Anthropic releases Claude Fable 5 with strict safeguards, sparking backlash from researchers

Anthropic says these topics are too dangerous to let its Fable 5 model talk about

Anthropic Tuesday publicly released Claude Fable 5, its first "Mythos-class" model that it says surpasses its previous frontier Opus models in overall capabilities. But the model's launch today comes with safeguards designed to prevent it from answering queries on topics like cybersecurity, biology, and chemistry, where the company has publicly worried about its potential impact to "uplift" malicious actors. Anthropic says Fable 5 operates on the "same underlying model" as Mythos 5, which is coming out of its monthslong "Mythos Preview" period today, but only for "a small group of cyberdefenders" judged trustworthy through the existing Project Glasswing. Unlike Mythos 5, though, the publicly accessible Fable 5 is designed to funnel queries on certain sensitive topics to the earlier Claude Opus 4.8 model and to warn the user when this is happening. Anthropic said it has tuned these safeguards to be "stricter than ideal," meaning the system may occasionally refuse "harmless requests" in a way that it acknowledges may be frustrating for regular users. But Anthropic says such false positives come up in less than five percent of all sessions in testing, and were worth it to avoid situations where Mythos could give malicious actors assistance in "causing serious harm that they couldn't have received from other sources." I can't let you do that, Dave Fable 5's topic-based safeguards are built around a system of classifiers designed to broadly detect banned prompt subjects as well as any potential jailbreak attempts. In over 1,000 hours of red-team testing with a bug bounty program, Anthropic says external teams failed to find any universal jailbreaks for Fable 5. The new model also resisted automated jailbreak attempts to a much larger degree than previous Claude Opus models, Anthropic said. The company said it is particularly worried about Mythos 5's ability to perform "agentic hacking," executing multi-part cyberattacks with much more facility than earlier models. But testing from the UK's AI Security Institute in recent months found that Mythos Preview performed similarly to OpenAI's GPT-5.5 on a suite of Capture the Flag challenges, suggesting Mythos' performance is not "a breakthrough specific to one model." Among the usual raft of fair-to-middling benchmark test improvements that Anthropic reports for Mythos 5 over previous frontier models, the company claims a significant jump in the model's capabilities on the cybersecurity-focused ExploitBench test. Mythos 5 scored a 78 percent on the benchmark's tests of vulnerable code exploits, a significant increase from the 40 percent score from Opus 4.8, and even the 69 percent score achieved by Mythos Preview. While earlier Anthropic models blocked bioweapons-related queries, that classifier now applies to all chemistry and biology-related queries in Fable 5. The company says it worries that "well-resourced malicious actors" could use even seemingly benign queries on these subjects to assist with "highly risky biological research" in a much more effective way than with previous models. Who can you trust? Anthropic seems to understand that making certain topics off-limits for Fable 5 is something of a double-edged sword. The company writes that "the same queries that are beneficial in the hands of cybersecurity professionals and biology researchers could be dangerous if available to malicious actors." That puts Anthropic in the somewhat awkward position of having to judge who is and is not trustworthy enough to have access to a model that it says has potentially dangerous capabilities. The company says it will be periodically expanding its existing Project Glasswing program "in consultation with the US government" to let in more cybersecurity professionals. That expansion will also include a new trusted access program for life sciences organizations that removes Fable 5's biology/chemistry safeguards while keeping cybersecurity safeguards in place. API and Enterprise users will be able to access the Fable 5 model at a cost of $10-per-million input tokens and $50-per-million output tokens starting today. Those prices are 67 to 100 percent higher than those for OpenAI's recent GPT-5.5, a difference that could be significant at a time when many users are balking at the high cost of frontier models. Anthropic's existing subscription plans will include access to Fable 5 through June 22, after which users will need to purchase "usage credits" to access the new model. Anthropic says it eventually hopes to restore Fable 5 access as a standard part of subscription plans once it has "sufficient capacity" to do so.

Cybersecurity researchers aren't happy about the guardrails on Anthropic's Fable

Anthropic released its latest model Fable on Tuesday, billing it as a public and limited version of its powerful and much-hyped cybersecurity model Mythos. But not everyone is happy with the restrictions, and a number of cybersecurity researchers and professionals have aired complaints online. "[Fable] rejects any request that could be tangentially cyber related. Even innocuous tasks like reading a blog post," said Valentina "Chompie" Palmiotti, a well-known security researcher who works at IBM X-Force. When a prompt triggers its guardrails, Fable pauses the chat and says that its "safety measures flagged this message for cybersecurity or biology topics." The guardrails were put in place to limit the risk that Fable could be used to develop malware or compromise software -- a longstanding concern within Anthropic. The restrictions on biology come from a similar concern around developing biological weapons. When the AI giant released Mythos in April, it restricted the model to a limited number of companies and organizations in what it called Project Glasswing, an effort to deploy the model to secure critical software and infrastructure. Last week, Anthropic expanded access to Mythos to hundreds of organizations in 15 countries. But despite the good intentions, many cybersecurity experts are still put off by the haphazard nature of the restrictions. Matt Suiche, a cybersecurity veteran, told TechCrunch that "if you ask it to write secure code, it assumes it is cybersecurity related work instead of software engineering best practices, and you get downgraded." Fable is programmed to fall back to Claude Opus 4.8 if it hits a guardrail. "It seems to be keyword based, so anything in the lexical field of 'cybersecurity' triggers the guardrails." "But it is understandable as we are still in the early days and they are still adapting their guardrails. I am sure they are going to evolve over time as Anthropic and other frontier model companies will collaborate more with the current new generation of cybersecurity companies," said Suiche, who is a member of the technical staff at Tolmo, an AI cybersecurity startup. "It's better to catch more people than not enough when you do such a release and to relax the guardrails over time." Another researcher griped on X that "even asking for a code review" triggers Fable's guardrails. Anthropic did not immediately respond to a request for comment. Apart from guardrails inside its models, Anthropic requires cybersecurity professionals to apply to the Cyber Verification Program. If they get approved, the applicants have fewer limitations on using Claude for cybersecurity work. OpenAI has a similar program called Trusted Access for Cyber.

Anthropic Walks Back Policy That Could Have 'Sabotaged' AI Researchers Using Claude

The company changed course after researchers spoke out against the policy, which would have covertly limited Claude's ability to develop competing AI models. Anthropic is backtracking on a policy that would have covertly limited competitors from using its new AI model, Claude Fable 5, to develop other AI models. The company changed course after the move received significant backlash from the AI research community. "We're changing Fable 5's safeguards for frontier LLM development to make them visible." Anthropic said in a statement to WIRED. "We made the wrong tradeoff and we apologize for not getting the balance right." Anthropic released Claude Fable 5, a version of its latest AI model with additional safety guardrails designed to prevent misuse, earlier this week. Some of the safeguards Anthropic decided on were unsurprising: The company said it would reroute users who asked questions about cybersecurity, biology, or chemistry to a less capable AI model to reduce the chances of someone using the advanced AI to carry out a cyberattack or build a bioweapon. But for researchers trying to use Claude Fable 5 for frontier AI development, Anthropic outlined a different approach. The firm would deliberately degrade the model's performance in ways that were invisible to the user. The move would effectively sabotage researchers trying to use Claude to train competing AI models, which Anthropic explicitly bans in its terms of service. Got a Tip?Are you a current or former Anthropic employee who wants to talk about what's happening? We'd like to hear from you. Using a nonwork phone or computer, contact the reporter securely on Signal at mzeff.88. Anthropic now says it's changing course, and that Claude Fable 5's safeguards for AI development will be visible to users. If the company suspects a user is trying to use Claude to build a highly capable AI it will alert them that it's either refusing the request, or rerouting the user to a less capable model. Anthropic reversed the policy after it received fierce backlash from the AI research community. Anthropic has already taken steps to limit competitors from using Claude to build closed and open source AI models, but critics say that quietly degrading the model's performance for certain users went a step too far. Claude's coding agent has become a favored tool among developers, including those working on open-source AI research projects, and researchers tell WIRED that the company's latest policy could have led to a troubling future in which only a handful of leading AI labs could perform advanced AI research. Dean Ball, a senior fellow at the Foundation for American Innovation and a former advisor to the White House on AI, wrote in a post on X that "degrading performance on ML research *without telling the user* is shockingly hostile and a terrible look." He continued in another post that the "secret sabotage" policy undermines Anthropic's overall stance, because it limits AI researchers from collaborating on AI safety. "It felt like Anthropic was saying to the public, 'We don't trust anybody else to do AI research. We are the only ones who have to do AI research," says Will Brown, research lead at the open source AI startup Prime Intellect. "It feels a bit like they're starting to pull the ladder up behind them." Brown said the policy would also have left developers in the dark about whether they were violating Anthropic's rules, since the company wouldn't alert them when its safeguards were triggered. He added that the restrictions could have had widespread consequences. For example, he pointed to the growing ecosystem of third-party evaluation firms that test frontier models for safety, performance, and reliability -- work that could have been hindered if Anthropic secretly degraded its model. Anthropic said it implemented the measures because Claude has become increasingly effective at accelerating AI research. In a recent blog post, the company said it is concerned that AI could improve its capabilities faster than society can adapt to them. Anthropic argued that it would be "good for the world to have the option to slow or temporarily pause frontier AI development to enable societal structures and alignment research to keep up." "These safeguards prevent foreign adversaries from using our most capable models in ways that pose severe safety risks. The US and its allies hold an edge in frontier chips and the highly optimized software that runs them at full potential," the company said in a statement to WIRED. "These safeguards ensure Claude isn't used to erode that advantage -- by optimizing chips developed by those adversaries, for example... In deciding whether to make them visible or invisible we faced a choice. A hidden safeguard is harder to probe and work around. This means the safeguards can be targeted much more narrowly." Anthropic says that because this safeguard around AI development is now visible, it needs to cast a wider net, meaning more benign requests may trigger its safeguards. The company says it's working to make its classifiers more precise as quickly as possible.

Anthropic's Claude Fable is a version of Mythos the public can access today

Anthropic is bringing its most powerful AI model to the general public for the first time, but it's doing it with guardrails. On Tuesday, the AI firm launched Claude Fable 5, the first publicly available version of its Mythos model. Anthropic says Fable 5 excels at software engineering, knowledge work, and vision, but it comes with hard safety limits. In high-risk areas like cybersecurity, biology, chemistry, and distillation, the model blocks responses and falls back to Claude Opus 4.8. Launched as a preview in April, Mythos was initially limited to a handful of partners due to cybersecurity concerns. Last week, Anthropic expanded access to hundreds of organizations across 15 countries, again focusing on organizations that manage critical infrastructure. Now, a version of that technology is available to anyone through Anthropic's Claude API and consumption-based Enterprise plans. Access on subscriptions will roll out in stages: through June 22, Fable 5 is be included in Pro, Max, Team, and seat-based Enterprise plans at no extra cost. On June 23, Anthropic will pull Fable 5 from those plans, requiring usage credits going forward, with plans to restore it as a standard subscription feature as soon as possible. Anthropic is also deploying a new version of Mythos, called Mythos 5, to organizations that have already been approved to access the advanced model. Fable's launch comes as Anthropic prepares to enter the public markets, alongside OpenAI and Elon Musk's SpaceX. It also follows the AI firm's plea urging major global AI labs to establish a coordinated brake pedal on frontier AI development. Anthropic warned that systems are advancing so rapidly that they may soon achieve recursive self-improvement (RSI), autonomously improving themselves without human intervention. Wary of what a Mythos-class model could do in the wrong hands, Anthropic says it stress-tested its classifiers with jailbreak attempts before releasing Fable 5. "Internally, we ran an external bug bounty that produced no universal jailbreaks in over 1,000 hours of testing. We then worked with external red-teaming orgs which also failed to find universal jailbreaks." That said, there could still be novel attacks remain possible. As a result, with the launch of Fable 5 and Mythos 5, Anthropic said it will require a 30-day retention on all traffic, even if enterprises previously had zero-retention agreements. Anthropic said it won't use the data for training, only to "defend against complex and novel attacks, including new jailbreaks," and "identify and reduce false positives." The policy could set an industry precedent in which access to increasingly powerful models comes with mandatory data retention policies framed as a safety measure. For those that continue to use the model, not every question will get a Fable 5 answer. Anthropic says the cases in which Fable has to defer to Opus 4.8 are rare, with early data showing at least 95% of Fable sessions running entirely on the model's own responses. In third-party testing, analytics company Hex said in a statement that Fable was the first to get a 90% on its core analytics benchmark of complex, long-running analytical tasks. "On the hardest questions, it shows strong judgement and attention to nuance," Hex said. Vibe-coding platform Base44 noted in a statement that Fable is better at "one-shotting full apps" and has excellent tool-calling. AI-powered workspace and agent platform Genspark said Fable beat every other model in its evaluations, and performed significantly better on tasks like UI design and game coding. Pricing for both Fable 5 and Mythos 5 is $10 per million input tokens and $50 per million output tokens, double the price of Opus 4.8. That price alone might serve as a deterrent for widespread use. Many enterprises are growing critical of AI costs after seeing the bills come in or blowing through their yearly AI budgets early. Advanced models like Opus 4.8 can exacerbate those issues, with advanced reasoning skills that can split a single request into multiple tasks. Anthropic said it expects demand for Fable 5 to be very high and difficult to predict. And indeed some, like shopping rewards platform Rakuten, might think the upside is worth the price point. "At the highest effort, Fable reflects on and validates its own work," Rakuten said in a statement. "For us, that's what makes highly autonomous operations possible -- the extra thinking pays for itself."

Anthropic's Claude Fable 5 Is the First Mythos-Level AI Model You Can Actually Use

AI enthusiasts have been aflutter since rumors first leaked in April about Anthropic's Claude Mythos Preview -- an AI model that the company claimed could sniff out so many cybersecurity vulnerabilities that, left unchecked, could break the internet. That's why Anthropic has slow-rolled its release. On Tuesday, Anthropic released the first model from the Mythos family that everyone can use: Claude Fable 5. In an extensive blog post, Anthropic spelled out why Fable 5 is the company's most advanced AI model while still maintaining safety guardrails. The company says the model offers improvements to help people with software engineering and knowledge work, and that it's better at understanding images and other nontext subjects. Apparently, it also beat Pokemon FireRed, something previous models had failed to do. Everyone wants to know what the Mythos family of models can do, but they require a lot of computing power to run. Because of that, Anthropic isn't making them available cheaply. In a couple of weeks, the company will start charging subscribers extra to use Fable 5. Is Claude Fable 5 safe? Anthropic said Claude Fable 5 is the safer version of Claude Mythos 5, and only trusted cybersecurity and software professionals should have access to Mythos 5 through Anthropic's Project Glasswing program. The Fable model has cybersecurity protections that Mythos doesn't have. "Fable 5 complied with zero harmful single-turn requests relating to planning a cyberattack, exploit development or defense evasion," the company said. "This held whether or not one of the requests used any of 30 different public jailbreak techniques." AI-enabled bioterror -- a horrifying phrase and even scarier reality -- has been a concern for the leading AI labs in recent months. Because of that, Anthropic decided not to let Fable 5 answer most questions about biology and chemistry, at least for now. Those questions will instead be routed to another model, Opus 4.8. Anthropic said it is working "in consultation" with the US government over plans to release Claude Mythos 5. Mythos was largely responsible for the White House's recent push to get AI companies to submit any new AI models for government review before their release. That plan isn't in effect as of now. In a change, Claude business users will now have to agree to a 30-day data retention policy. Anthropic said the data will be used to help the company defend against future cyberattacks and AI misuse, not to train new AI models. How to use Claude Fable 5 Fable 5 is available now for Claude subscribers, but its rollout will be done "more conservatively, in stages," Anthropic said in a blog post. The model will be available for paying subscribers -- on the Pro, Max, Team and enterprise plans -- from now until Monday, June 22. The next day, on Tuesday, June 23, Fable 5 will be removed from the subscriber options list. You'll still be able to use Fable 5 after the initial release period, but you'll need to burn usage credits. These are special "pay-as-you-go" credits that let you keep using Claude even after you've reached your limit. You may already have usage credits included in your subscription, especially if you're on an enterprise plan. But if you don't have any usage credits or exceed your allotted amount, you will see extra fees on your next bill. Developers can use Fable 5 now in the Claude API, priced at $10 per million input tokens and $50 per million output tokens. Anthropic's unusual rollout plan is likely due to capacity concerns. Fable 5 uses twice as much as Anthropic's Opus line of models, according to a pop-up on Claude when you select Fable 5. The company says it anticipates that demand will be high. So to accommodate that demand and keep Claude online, the initial availability window is only a couple of weeks. Anthropic said it aims to eventually make Fable 5 available as part of its paid plans, but there's no set timeline on when, or if, that may happen. For more, check out Apple Intelligence's WWDC makeover and the best AI chatbots.

Anthropic Offers Mythos Upgrade for Cyber Partners and a 'Safe' Version for the Rest of You

Anthropic released two new AI models called Claude Fable 5 and Claude Mythos 5 on Tuesday, which the company says have greater capabilities than the Mythos Preview model it released in April to a limited set of tech industry partners. Anthropic has said the initial, limited release stemmed from concerns that the model's capabilities could be exploited by bad actors to develop hacking tools that could catch defenders off guard. Anthropic is currently only releasing Claude Mythos 5 to a limited set of industry partners, many of which received access to Mythos Preview, and the company says it is collaborating with the US government on the rollout. Claude Fable 5, which is being publicly released, uses the same underlying model as Mythos 5, but will have "guardrails" in place at launch, the company said Tuesday, that will block the model from answering many user questions related to cybersecurity, biology, and chemistry. These requests will instead be re-routed to an older AI model, Claude Opus 4.8. If Anthropic suspects a user is trying to conduct distillation -- training a smaller AI model off a larger AI model's responses -- on Claude Fable 5, those requests will also be rerouted to Claude Opus 4.8, the company says. In an interview with WIRED, Anthropic's head of Product Management, Diane Penn, says that the company has been grappling with the question of how to handle Mythos's software vulnerability-discovery abilities and other advanced capabilities since before its April release, but that testing and user input since then helped to hone the strategy. "We're trying to make improvements in a way that's beneficial, even if we don't have the perfect [solution] for every use case to start," Penn says. "Out of all the different approaches, this emerged as the most viable and the best one. We just ended up feeling like this was the best product choice for users to get the maximum value out of Fable 5." For now, Penn says that the protective mechanism is built to err on the side of caution, meaning some user queries may be routed to the less capable AI model even if they're benign. Over time, Anthropic hopes to make its classifiers more precise, but Penn says this was the only safe way the company could release the model broadly at this time. The company said on Tuesday that in addition to offering Claude Mythos 5 to Project Glasswing partners, it is also giving access to "select biology researchers." Additionally, Anthropic noted in its blog post about Tuesday's launch that it is providing unrestricted versions to these small groups of customers "until our trusted access program is available," hinting at future plans to expand access even more. Since the Mythos launch in April, Anthropic has repeatedly emphasized that eventually its competitors in both the private and even open weight spaces will inevitably also offer models with Mythos-level capabilities. The ability for Claude Mythos and other new AI models to design hacking tools that can find and exploit vulnerabilities in both new and legacy software has forced tech companies and governments around the world to secure their software defenses before AI models of this level are made broadly available to attackers. Anthropic first released Mythos to industry partners under a consortium called Project Glasswing, with the idea that this could give members a head start in preparing their own systems and weighing global solutions to the threat before a broader release. Anthropic wrote in an update about Project Glasswing last week: "We're working as quickly as we can to safely release Mythos-level capabilities in general access. To do so, we'll need highly robust safeguards that prevent the model's cyber capabilities from being misused -- safeguards that we (and, to our knowledge, all other AI developers) have yet to develop." Anthropic says Claude Fable 5 -- named after the literary form, much like the company's existing Haiku, Sonnet, and Opus models -- offers increased performance on software engineering and tasks that require visual understanding. But that added performance comes at a price. Claude Fable 5 and Claude Mythos 5 will cost developers $10 per million input tokens and $50 per million output tokens -- twice as much as Anthropic's publicly available AI models, but cheaper than Mythos Preview. The neutered release of Claude Fable 5 hints at Anthropic's business tension of wanting to release a Mythos-class AI model for general use before the tech industry has resolved the cybersecurity concerns of these models. In April, OpenAI also privately launched a model that it said has advanced cybersecurity capabilities and convened a working group similar to Project Glasswing. Both OpenAI and Anthropic have confidentially filed for IPOs, and are racing to impress prospective investors before they become public companies as soon as this year. Even as an interim solution, though, it remains to be seen how resistant Claude Fable 5's safeguards are in the wild. Anthropic says in more than 1,000 hours of red-teaming, its testers found no universal jailbreaks for the model. Still, fears about the ability to develop adequate protections underpinned the company's original justification for why it did not release Mythos-class models to the public in April, and these fears have seemingly persisted.

Anthropic backpedals on Fable safety measure

Anthropic has apologized for stealthily throttling its new AI model, Claude Fable 5, with hidden guardrails that undermine both researchers and rivals using it to develop competing systems. The company says it is reversing course and will be more transparent about when the restrictions kick in, even if that means Fable refuses more queries. Fable is the first widely available model in Anthropic's Mythos class of AI systems, a group the company has spent months warning are too dangerous for public release. Anthropic says it has addressed some of those risks by launching Fable with safeguards that prevent it from responding to certain "high-risk" queries. One of the areas Anthropic said it would restrict Fable's responses is distillation, a technique for training smaller AI models using the outputs of larger ones. In Fable's system card -- a public document AI developers release to explain how a system works -- Anthropic said it would handle queries it believed were distillation attempts by altering and degrading the model's answers directly. Users would not be notified that they had triggered the safety measure or informed that the responses had been changed. Anthropic said it is now changing its approach to distillation: Queries will now fall back to Claude Opus 4.8, Anthropic's previous flagship model, the company said in a post on X. Anthropic will prominently tell users too: "You will see this every time it happens." This is similar to how Fable handles queries in other high-risk areas. When safety features are triggered in areas like biology, chemistry, and cybersecurity, queries are routed through Opus 4.8 unless they are blocked outright under the company's broader safety rules, such as those covering drugs, weapons, or other prohibited content. In some cases, notably biology, the safeguards have been calibrated so broadly that Fable is practically unusable for even basic queries, something Anthropic acknowledged in a comment to The Verge. "Visible safeguards can be probed, so they have to be robust, which takes time to get right," Anthropic wrote. "Invisible safeguards can be targeted more narrowly, allowing us to ship quickly with very few false positives. We went with invisible safeguards for this reason -- and that was the wrong tradeoff. You should have visibility into the safeguards we have in place, and why. We're sorry for not getting the balance right." The change follows intense backlash from the AI research community over Anthropic's decision to silently limit users suspected of trying to distill Fable into competing models -- a safeguard critics warned could also affect third parties trying to evaluate the frontier model. In the system card, Anthropic said newer models' ability to accelerate AI development justified targeting those requests, noting that "using Claude to develop competing models already violates our Terms of Service." Anthropic has previously accused Chinese rivals like DeepSeek of unfairly distilling its models on an "industrial" scale.

How Fable 5 And Mythos 5 Change AI Security, Data Retention, And Vendor Risk

Anthropic's Fable 5 and Mythos 5 is the most 2026 product launch you'll read this year. The same model can find nation-state zero days, design novel drug candidates, and play FireRed on a Gameboy Advance with nothing but screenshots. And for the gaming fans out there: yes, we got Fable 5 before Fable 4. These three examples also exhibit the characteristics that matter in these models: long time horizon tasks, self-correcting operations, and autonomous reasoning. Anthropic released all three of these on Tuesday under two names: Claude Fable 5 for the public and Claude Mythos 5 for a vetted few. For Forrester clients, we've prepared an update on how to discuss this release with your Board of Directors and Senior Leaders here. What's Understood And Expected Anthropic was always going to ship Mythos. In April, the company introduced Mythos Preview as a model too dangerous to release, gated to roughly 50 Project Glasswing partners. The partnership expanded in June to 150. Its "eventual" release arrived a few weeks later. A frontier lab on the brink of IPO sitting on its most capable model indefinitely was never going to happen. Fable 5 is that model with guardrails; Mythos 5 is that model with the guardrails selectively lifted in some use cases for defenders. Fable 5 is state-of-the-art on nearly every capability benchmark, and both models list at less than half the price of Mythos Preview. These models are still expensive, just less expensive than the Mythos preview version. And the harsh truth remains: as token prices come down, token usage goes up, meaning the total amount spent increases even as the per unit costs drops. Mythos requires approved access and a hefty wallet to afford it. The divide between haves and have nots in cybersecurity just widened, ushering in an era of aggressive model cost optimization and hard tradeoffs on capability versus cost. What To Know About Two Big Changes The Safeguards Are A Control You Depend On...But Don't Operate Fable 5 and Mythos 5 are one model with a safety switch. For cybersecurity, biology, and chemistry topics, Fable 5 can block the request and route the query to Opus 4.8 instead. Users are informed the fallback took place. Anthropic says the fallback triggers in under 5% of sessions, is tuned conservatively so it will sometimes catch harmless requests, and that testing surfaced no universal jailbreaks. Mythos 5 is a gated release to select members of Project Glasswing, but it removes some of the safeguards that Fable 5 ships with. This aligns the model release with our first in market AEGIS Framework principle of Least Agency tuning and guardrails at the model provider layer. Your enterprise does not administer this guardrail. The guardrails' scope, sensitivity, and reliability are set by Anthropic. defining what acceptable risk looks like for global R&D and security operations teams, without any real public governance structure. Enterprise-specific runtime security and guardrails for applications and users working with these models is still necessary for anything your organization wants to detect or prevent. The Data Retention Policy Change Is A Huge Adjustment For Enterprises There's another significant change coming with Fable 5 and Mythos 5. Anthropic now requires 30-day retention on all traffic (prompts and completions) across both its own surfaces and third-party platforms - a requirement that overrides existing zero-retention agreements. If your enterprise negotiated a zero-retention DPA, using a Mythos-class model voids it for that traffic. There is no opt-out. Consumer subscriptions already included a retention period, so this doesn't change for Pro and Max subscribers, which is why we continue to advise caution about what data consumers put into these tools. For third-party risk concerns, Anthropic says the data will not train new Claude models and will not be used for any non-safety purpose; that it logs all human access to the retained data; and that it deletes the data after 30 days in almost all cases. Anthropic published a whitepaper in its trust center about the use cases and safeguards in place to protect enterprise data, and highlights when retained data is kept beyond the 30-day window. The stated purpose is defensive: catching novel attacks, multi-request abuse, and new jailbreaks, and reducing false positives in the safeguard layer. The 30-day window lines up with a White House executive order that set a voluntary framework for AI companies to share frontier models with the government ahead of public release. So "safety monitoring" and "potential government visibility" are now adjacent concepts. What To Do Fable 5 is the most capable model the public can touch today and the benchmarks are laudable. The changes to data retention, silent model downgrades, and premium pricing all have immediate impact on cybersecurity teams. Accept that availability may need to suffer to preserve confidentiality. That's a tradeoff that will be difficult for IT and Development teams that consider uptime and "5 9's" as sacred, but it's the reality of the world we now live in. Deploying untested patches or virtual patching to prevent a potential exploit is worth the tradeoff in potential downtime -- which major banks are preparing for now Giving customers partial bill credits for outages is far less expensive than a decade of litigation and fines from regulators from a data breach. This isn't just a technology decision; it's a behavioral and cultural shift that will make many in the organization uncomfortable. Frontier AI makes this a necessary evil. Patch fatigue will morph into triage fatigue as tired teams of defenders tap out from the never-ending onslaught of issues that must be tracked and remediated from these model releases. Take the following steps to counter this before it overburdens your teams and slows your decision making: Open-source Maintainers Still Need Help These disclosures require investigation, triage, and, when applicable, remediation due to expanded Mythos 5 access. To assist, qualifying open-source maintainers can sign up for free access to the Max20x plan for six months, which offers higher usage limits. However, this assumes that maintainers have the time and capacity to prioritize addressing the surge in reported vulnerabilities. Anthropic, like most AI model providers, was built on open-source software and continues to benefit from the open-source ecosystem. With another blockbuster revenue quarter projected, sustained funding to critical open-source projects would demonstrate a commitment to the community that Mythos is likely to disrupt. Now is the time for ruthless prioritization of the open-source software you use: well-maintained, communicative, and security-oriented deployments. In a worst-case scenario, for less-maintained but critical open-source software, it can make sense to fork it and take on that maintenance yourself. The First Document Defenders Will Create Whether it's your pen testers or your SOC Analysts, the first document almost every team will share is not going to be "best prompts for Fable 5". It will be "Prompts that bypass fallbacks in Fable 5" that explains how to get around safeguards to obtain Mythos 5 like capabilities and bypass the conservatively tuned safeguards that impact up to 5% of sessions. Anthropic attests to 1000s of hours of testing to prevent jailbreaking Fable 5, but motivated and creative security pros always find a way. Your Old Processes And Procedures Don't Work Anymore Anthropic benchmarks showcase Fable 5 working across long time horizon sessions over days, delegating to subagents, checking its own work, and recursively improving its own code. Your assumptions of change control, application security gates, testing rules, and two-person approval workflows were not built for self-improving autonomous software development that finishes testing in the time it takes you to schedule meetings between two busy people. This requires process reinvention across multiple domains in cybersecurity, and a shift in your risk appetite that differs from what your organization is used to. Establish Agentic Development Security Practices As developers adopt Fable 5 and other advanced coding agents, the increased volume, speed, and complexity of releases will surpass what traditional application security testing tools were built to address. Organizations must invest in Agentic Development Security (ADS) tools designed to prevent insecure code generation, AI selection of hallucinated, outdated, or vulnerable third-party components, and agents leaking sensitive data and secrets. Coding agents introduce their own software supply chain, which encompasses MCP servers, skills, configuration files, extensions, and models, all of which expand the attack surface and inherit developer access to sensitive data, cloud credentials, source code repositories, and productivity tools, along with permissions to read, write, and execute destructive commands. Fable 5 incorporates safeguards to prevent cybersecurity misuse, like other coding models, It does not, however, offer sufficient visibility into the supply chain risks it inherently carries. Organizations need to ensure coding agents follow least agency principles and operate in sandboxed environments properly guard-railed by ADS tooling. Prepare For Frontier Level Capabilities In Regular SaaS Vendors With Fable 5 now generally available, any vendor in your ecosystem can turn on a Mythos‑class model overnight, even before your organization ever "adopts" it. Ostensibly, ordinary vendors are running frontier level capabilities that you never assessed, governed by AI safety practices you never vetted, and adding another layer of complexity to third-party risk management (TPRM). Vendors using Fable 5 should move into a different criticality tier, face AI‑specific risk assessments, and be covered by explicit AI, data, and safety obligations in your contracts, with particular emphasis on tightening legacy suppliers still operating on flimsy pre‑AI era paper. Get Ahead Of Security Tech Contract Renewals Given how quickly security tools and platforms are becoming orchestration and data-wrapper layers around Mythos‑class models, use this next renewal cycle to reset how you buy, evaluate, and govern "AI‑powered" security tech. Start by inventorying which vendors already AI in their tools and flag those for accelerated, deeper review. As part of the renewal process. demand a concrete 12-24‑month roadmap from each strategic vendor for Mythos‑class model adoption, new use cases, pricing impact, and product security. For new procurements, require precise disclosure of foundation models, data flows, guardrails, and how they plan to balance capability against cost on your behalf. Harden contracts and DPAs around data handling, logging, and retention when invoking third-party models, and guardrails for validating AI outputs before anything is allowed to drive automated actions. Connect With Us Forrester clients with questions related to this can connect with us through an inquiry or guidance session.

Anthropic's new Claude Fable 5 is a nerfed Mythos with guardrails attached

Follow ZDNET: Add us as a preferred source on Google. ZDNET's key takeaways * Anthropic is releasing Claude Fable 5 for general users. * Fable 5 uses Mythos-class power with safety controls. * Pricing is about twice that of Claude Opus 4.8. Anthropic has announced a defanged version of its fabled (and highly restricted) Mythos large language model. Called Claude Fable 5, the company describes the new AI as "a Mythos-class model made safe for general use." Mythos was introduced back in April to great fanfare as a model capable of finding vulnerabilities in code that neither experienced developers nor other AIs could find. Also: Apple, Google, and Microsoft join Anthropic's Project Glasswing to defend world's most critical software Offered as the central component of a Manhattan Project-like team effort called Project Glasswing, Mythos was considered far too dangerous to be allowed to fall into the wrong hands. As such, the model has been made available only to Glasswing partners, including Amazon Web Services, Anthropic, Apple, Broadcom, Cisco, CrowdStrike, Google, JPMorganChase, the Linux Foundation, Microsoft, Nvidia, and Palo Alto Networks. Up until now, Mythos was considered a preview product. Now, Anthropic is launching Claude Mythos 5, which will be available to all users with access to Mythos Preview. The company said, "We plan to expand access over time through a more systematic trusted-access program." No significant details were provided about any difference between Mythos Preview and Mythos 5, except that the latter appears to be the post-beta version of Mythos. Also: Apple's new Siri AI comes with hidden costs that power users should know of Anthropic also released Fable 5. The company described this technology as the same underlying model as Mythos, but with safeguards. Those safeguards block responses in specific high-risk areas of cybersecurity and biology. Yeah, the "biology" reference raised red flags. Have they seen biological weapon prompts or responses in their logs? I've asked Anthropic. I highly doubt they'll answer. Also: Anthropic's new Claude Security tool scans your codebase for flaws - and helps you decide what to fix first Interestingly, if a Fable 5 prompt veers into one of those high-risk areas, the model drops down to Opus 4.8, which has its own restrictions. Ever since version Opus 4.7, Anthropic blocks "Activities that are almost always used maliciously and have little to no legitimate defensive application, such as mass data exfiltration or ransomware code development." Professionals with a security clearance from Anthropic can use Opus 4.7 and 4.8 to perform blocked security activities in the course of doing their jobs. Disclosure: I am an authorized member of Anthropic's Cyber Verification Program, so I have access to these capabilities as part of my cyberwarfare, cyberdefense, and counterterrorism work. It's not yet clear whether those of us certified through the Cyber Verification Program can perform blocked queries with Fable 5. Also: I used ChatGPT to build a free PDF editor because I didn't trust it to change my files - it's glorious Anthropic seems confident that Fable can't be utilized for nefarious purposes. They said, "Early data shows at least 95% of Fable sessions run entirely on Fable's own responses, with no fallback. We extensively red-teamed our classifiers to test their robustness against jailbreaks. Internally, we ran an external bug bounty that produced no universal jailbreaks in over 1,000 hours of testing. We then worked with external red-teaming orgs which also failed to find universal jailbreaks." From the announcement, we don't have much information to share with you about Fable 5. However, Anthropic did share some customer comments about the release. An unnamed representative of vibe-coding platform maker Base44 said, "Fable is much deeper and better at one-shotting full apps, and its tool calling is excellent." Also: I stopped using ChatGPT for everything: These AI models beat it at research, coding, and more Genspark provides an all-in-one AI workspace. A representative from that company said, "Fable came out #1 on our evals, winning head-to-head against every model we tested. It was significantly stronger on the hardest tasks in the set -- UI design and game coding." A representative from e-commerce marketplace Rakuten said, "At the highest effort, Fable reflects on and validates its own work. For us, that's what makes highly autonomous operations possible -- the extra thinking pays for itself." Speaking of payment, pricing for Fable 5 and the new Mythos 5 release is $10 per million input tokens and $50 per million output tokens. That's about twice the price of Claude Opus 4.8. Anthropic has a fairly unique rollout plan for Fable 5, based on their expectation of high demand. Here's what the company said: * From today through June 22, Fable 5 is included on Pro, Max, Team, and seat-based Enterprise plans at no extra cost. * On June 23, we'll remove Fable 5 from those plans. Using it after that will require usage credits. * After this point-we aim to restore Fable 5 as a standard part of subscription plans. We intend to do this as quickly as we can. Anthropic hasn't provided any deep explanation of its naming choices for Mythos and Fable. According to Britannica, a mythos is a "complex body of sacred narratives, traditional stories, or belief systems that explain the origins of the world and cultural values." By contrast, a fable is a "short, fictional story designed to teach a specific moral lesson." Read into that what you wish. I'm sure we'll learn more about both of these releases. If my Max plan can host Fable 5, I'll try and test it out with some coding challenges when it becomes available. So stay tuned. Even at twice the price of Opus, would you use Claude Fable 5 for coding work if it gave you Mythos-level power with built-in safety fallbacks? Let us know in the comments below. You can follow my day-to-day project updates on social media. Be sure to subscribe to my weekly update newsletter, and follow me on Twitter/X at @DavidGewirtz, on Facebook at Facebook.com/DavidGewirtz, on Instagram at Instagram.com/DavidGewirtz, on Bluesky at @DavidGewirtz.com, and on YouTube at YouTube.com/DavidGewirtzTV.

Anthropic releases Claude Fable, a version of Mythos, days after warning AI is becoming too dangerous

Anthropic is bringing its most powerful AI model to the general public for the first time, but it's doing it with guardrails. On Tuesday, the AI firm launched Claude Fable 5, the first publicly available version of its Mythos model. Anthropic says Fable 5 excels at software engineering, knowledge work, and vision, but it comes with hard safety limits. In high-risk areas like cybersecurity, biology, chemistry, and distillation, the model blocks responses and falls back to Claude Opus 4.8. Launched as a preview in April, Mythos was initially limited to a handful of partners due to cybersecurity concerns. Last week, Anthropic expanded access to hundreds of organizations across 15 countries, again focusing on organizations that manage critical infrastructure. Now a version of that technology is available to anyone through Anthropic's Claude API and consumption-based Enterprise plans. Access on subscriptions will roll out in stages: Through June 22, Fable 5 will be included in Pro, Max, Team, and seat-based Enterprise plans at no extra cost. On June 23, Anthropic will pull Fable 5 from those plans, requiring usage credits going forward, with plans to restore it as a standard subscription feature as soon as possible. Anthropic is also deploying a new version of Mythos, called Mythos 5, to organizations that have already been approved to access the advanced model. Fable's launch comes as Anthropic prepares to enter the public markets, alongside OpenAI and Elon Musk's SpaceX. It also follows the AI firm's plea urging major global AI labs to establish a coordinated brake pedal on frontier AI development. Anthropic warned that systems are advancing so rapidly that they may soon achieve recursive self-improvement (RSI), autonomously improving themselves without human intervention. Wary of what a Mythos-class model could do in the wrong hands, Anthropic says it stress-tested its classifiers with jailbreak attempts before releasing Fable 5. "Internally, we ran an external bug bounty that produced no universal jailbreaks in over 1,000 hours of testing. We then worked with external red-teaming orgs which also failed to find universal jailbreaks." That said, there could still be novel attacks. As a result, with the launch of Fable 5 and Mythos 5, Anthropic said it will require a 30-day retention on all traffic, even if enterprises previously had zero-retention agreements. The company said it won't use the data for training and will use it only to "defend against complex and novel attacks, including new jailbreaks," and "identify and reduce false positives." The policy could set an industry precedent in which access to increasingly powerful models comes with mandatory data-retention policies framed as a safety measure. For those who continue to use the model, not every question will get a Fable 5 answer. Anthropic says the cases in which Fable has to defer to Opus 4.8 are rare, with early data showing at least 95% of Fable sessions running entirely on the model's own responses. In third-party testing, analytics company Hex said in a statement that Fable was the first to get a 90% on its core analytics benchmark of complex, long-running analytical tasks. "On the hardest questions, it shows strong judgement and attention to nuance," Hex said. Vibe-coding platform Base44 noted in a statement that Fable is better at "one-shotting full apps" and has excellent tool-calling. AI-powered workspace and agent platform Genspark said Fable beat every other model in its evaluations and performed significantly better on tasks like UI design and game coding. Pricing for both Fable 5 and Mythos 5 is $10 per million input tokens and $50 per million output tokens, double the price of Opus 4.8. That price alone might serve as a deterrent for widespread use. Many enterprises are growing critical of AI costs after seeing the bills come in or blowing through their yearly AI budgets early. Advanced models like Opus 4.8 can exacerbate those issues, with advanced reasoning skills that can split a single request into multiple tasks. Anthropic said it expects demand for Fable 5 to be very high and difficult to predict. And indeed some, like shopping rewards platform Rakuten, might think the upside is worth the price point. "At the highest effort, Fable reflects on and validates its own work," Rakuten said in a statement. "For us, that's what makes highly autonomous operations possible -- the extra thinking pays for itself."

Claude Fable 5 brings Mythos to the masses -- Anthropic's next frontier model is 'state-of-the-art on nearly all tested benchmarks'

Queries regarding cybersecurity, biology and chemistry, and distillation will be redirected to the prior-gen Opus 4.8, however Ever since Anthropic first made the earth-shaking disclosure of its incredibly capable Claude Mythos AI model back in April and the steps it was taking toward a safe release of that product, the AI public has been waiting with bated breath to get its hands on its capabilities to give them a spin. Now, with today's release of Claude Fable 5, Anthropic says it finally has a model of this class that's "safe for general use." Anthropic is also releasing the unrestricted Mythos 5 model to members of its Project Glasswing program for use in sensitive cybersecurity and biology contexts. As with every new cutting-edge model from frontier labs, Anthropic has a selection of benchmarks of Fable 5's performance across a range of widely accepted tests that highlight its state-of-the-art-ness, but those numbers aren't as interesting as the specific use cases that the company highlights for this level of capability. For example, the company highlights how Stripe was reportedly able to perform the migration of a 50-million-line Ruby codebase in a single day that would have otherwise required two months of team effort had it been performed by hand. That sort of task compression on a job of such massive scope illustrates how Fable 5 and Mythos 5 "can work autonomously for longer than any previous Claude models," in addition to their high overall level of capability. For vision tasks, Anthropic simply says that Fable 5 is "the new state-of-the-art model." Among other accomplishments, the company says it was able to play through Pokemon FireRed in its entirety using only a "minimal, vision-only harness." Past models apparently struggled to complete this task even with the ability to seek outside help via tool-calling. Wharton School professor and AI blogger Ethan Mollick also has vivid examples of what Fable can do. Among other tasks, he describes how he gave Fable a 19-page spec document for the development of a categorization and analysis tool for unstructured survey answers. He describes how the model worked for "nine and a half hours" to generate an "extremely sophisticated" tool that "researchers have needed for years but was never profitable to create." In order to keep Mythos-level capabilities out of the hands of malicious actors, Anthropic says it will be redirecting queries on certain topics, namely "cybersecurity, biology and chemistry, or distillation," to the last-gen Claude Opus 4.8 instead. The company says that users will be informed whenever this redirection occurs, and that it should trigger in "less than 5%" of interactions with the model. Mollick, however, says these limitations trip "at the faintest hint of a security problem," suggesting that even the well-intentioned won't be able to use Fable 5 to bolster the security of their code bases. According to its model card, Fable 5 will also be nerfed in instances where a user attempts to use it to further cutting-edge AI or ML research, which is likely not only motivated by the company's recent concerns around AI self-improvement but also likely to be motivated by competitive concerns with other labs and with geopolitical actors. Within those guardrails, Anthropic says Fable 5 is available everywhere today, and that access to it will be billed at $10 per million input tokens and $50 per million output tokens via the company's API. That's twice the cost of the now-last-gen Opus 4.8 and just over 3X the cost of Sonnet 4.6. Those on Anthropic's Pro, Max, Team, and Enterprise subscription plans will have a short gratis access window to Fable 5, from now until June 22. After that point, access to the model will require paying for usage credits. The company says it will restore access to Fable 5 through these plans "as quickly as we can" when it has the compute capacity to do so. The world hasn't ended as of this writing now that a public Mythos-class model is available. It'll surely be interesting to see what problems people are able (and aren't able) to tackle with an AI model of this level of capability -- and with this degree of restriction around its capabilities. The only thing we can say for certain is that things are only going to get weirder from here. Follow Tom's Hardware on Google News, or add us as a preferred source, to get our latest news, analysis, & reviews in your feeds.

It blocked us at 'hello!' Anthropic Fable 5 refusing innocuous prompts

Anthropic's newly released Claude Fable 5 generative AI model is trying so hard to be safe that it's hurting its own userbase. Customers attempting to use the AI knowledge regurgitator are reporting that the model is refusing to answer harmless questions, an issue that has annoyed security researchers following past model releases. Anthropic warned that it had tuned Fable 5's guardrails conservatively: "they'll sometimes catch harmless requests, though they trigger, on average, in less than five percent of sessions," the company said, promising to "reduce false positives as quickly as we can." The company did not immediately respond to a request to quantify model refusals. So it's unclear whether the actual false positive rate is greater or less than five percent. But with an estimated 18 to 30 million users worldwide, even a small percentage of thwarted users makes a racket. Mike Famulare, principal research scientist at the Institute for Disease Modeling, part of the Global Health Division of the Gates Foundation, reports (#66657) that Claude Fable 5 balks at inputs like "Hello." "In Claude Code, Fable 5's input safety classifier emits a model_refusal_fallback (silent switch to Opus 4.8) on the first turn of essentially every session on my account -- including a session whose only user input is the word hello!. No repo content, no tool calls, and no file reads are in context when it fires." He is not the only frustrated customer. Many other bug reports have been filed in Anthropic's Claude Code GitHub repo since Fable 5 debuted. These include: [Bug] Fable 5 model safety filters causing false positives on benign messages #66587; Fable 5 refuses to assist with 'Application Security Architect resume' editing #66655; and [Feature Request] Allow Fable 5 usage for non-research lab management systems #67062, among others. On social outrage site X.com, Derya Unutmaz, an immunologist and professor at the Jackson Laboratory for Genomic Medicine, notes, "The word 'cancer' is flagged as a biosecurity risk by Claude Fable 5!" Similar complaints show up in Reddit threads. Fable 5 is unusual because Anthropic has chosen to conceal safety interventions that try to block rival frontier model development. The classifiers designed to catch cybersecurity, biology and chemistry, and distillation attempts fall back on the latest Claude Opus model and the user gets notified. But the counter-competition surveillance, per the company's system card [PDF], "will limit effectiveness through methods such as prompt modification, steering vectors, or parameter-efficient fine-tuning (PEFT)." "Prompt modification" without notice is functionally a man-in-the-middle attack, though one that Anthropic estimates "will impact ~0.03 percent of traffic, concentrated in fewer than 0.1 percent of organizations." As developer Clay Merritt fumes, "Anthropic's Fable 5 silently sabotages its answers when it detects AI/ML work. No refusal. No notice. Purposeful degradation invisible to the user." Anthropic expects cyber defenders and critical infrastructure providers to use its Claude Mythos 5 model, which shares the underlying model of Fable 5 but without the same safeguards. Doing so, however, requires participating in the company's Project Glasswing program or the trusted access program that's being rolled out for select biology researchers. Devon (last name withheld by request), founder of Abliteration.ai, a service that assists with model abliteration (guardrail removal), told The Register in a phone interview that while there's some degree of fearmongering and marketing hype coming from the big AI labs, it's also fair to say that there are legitimate concerns about how frontier models get used. "Anthropic's making a big bet on their brand that people will trust their brand so much they'll just deal with [refusals]," he said. "But in the long term, people are not just going to accept these companies that centralize control over their lives and what they can have information about." ®

Anthropic rolls out Claude Fable 5, but it's available for a limited time

Anthropic has begun rolling out a new model called "Fable," which is based on the same underlying model as Mythos, its most powerful AI model class. Anthropic previously said that it developed a model called "Mythos," which is a state-of-the-art model that poses security risks to companies around the world. At that time, Anthropic noted that Mythos was powerful enough to potentially help bad actors attack public and private software. "The advantage will belong to the side that can get the most out of these tools," Anthropic warned in April when it announced the Mythos model. "In the short term, this could be attackers, if frontier labs aren't careful about how they release these models. In the long term, we expect it will be defenders who will more efficiently direct resources and use these models to fix bugs before new code ever ships." In other words, it could have been abused to find and exploit vulnerabilities in apps like Firefox. Because of those risks, Anthropic decided to limit access to models like Mythos and offer them only to cybersecurity experts and trusted companies. Now, Anthropic says it has developed strong guardrails for the same model class, which means these powerful AI models can no longer be easily exploited by bad actors. As a result, it's launching a safer version called "Fable 5." According to Anthropic, this model has strict safeguards in place that will block or divert sensitive queries, like those involving offensive cybersecurity, biology, or chemistry, to its previous model, Opus 4.8. Claude Mythos 5 is the unrestricted version of that same model, with those safeguards lifted. Because of the risks involved, it is only available to a highly vetted group of trusted partners, such as government cyberdefenders and specific life sciences researchers. Fable 5 is free for a limited time, and it consumes tokens faster than any other model. Anthropic says Fable 5 is an expensive model because it requires a lot of compute, which means the company cannot afford to make it available as easily as Opus 4.8 or its previous models. However, until June 22, Anthropic says Fable 5 will be offered to all Pro, Max, and Enterprise customers, but after the window expires, it'll switch to usage-based pricing. In our tests, BleepingComputer observed that Fable 5 uses a massive amount of tokens in a span of minutes. I particularly noticed this behaviour when I used Workflow, a new execution system that allows Claude to break complex prompts into smaller tasks and spin up parallel subagents to implement them. Claude Fable 5 exhausted my $100 Max subscription's daily usage, which was at zero when I started using it, in just 9 minutes This doesn't happen when you casually interact with Claude Fable 5, but if you switch to Workflow mode and change model thinking to high, you're going to consume all your tokens in minutes. However, even if you don't use Fable 5 with workflow in xhigh effort, you're still going to consume it 2 times faster than Opus model. This explains why Anthropic is hesitating to unlock Fable 5 in the same capacity as Opus and other models, but that could change in the coming weeks, as the company is known for nerfing its models and increasing the capacity later.

Anthropic Releases Claude Fable 5, Its Most Powerful AI Yet, With Cyber Safeguards

On June 9, Anthropic released Claude Fable 5, the most capable model it has ever made, generally available. It also did something unusual: it shipped one model as two products, split not by capability but by a layer of safety classifiers. Fable 5 goes to the public. Its twin, Claude Mythos 5, the same underlying model with the cyber safeguards lifted, stays locked to a vetted group of cyber defenders and critical infrastructure operators. Anthropic calls Mythos 5 the strongest cybersecurity model in the world. The practical difference is this: Fable 5 routes flagged cyber, biology, chemistry, and distillation requests to the weaker Claude Opus 4.8, while Mythos 5 keeps the cyber capabilities available for vetted users. Both models cost $10 per million input tokens and $50 per million output tokens, less than half the price of the earlier Mythos Preview, and Fable 5 is available through the Claude API now. It is included on Pro, Max, Team, and seat-based Enterprise plans at no extra cost through June 22, then moves to usage credits. How Fable 5's cyber classifiers work The split exists because Mythos-class models find and exploit software vulnerabilities well enough that, in Anthropic's framing, handing that capability to the general public without controls would give attackers serious uplift. The mechanism is a set of classifiers: separate AI systems that watch for misuse and jailbreak attempts. When a request trips one, Fable 5 does not refuse. The response is handed to Opus 4.8, and the user is told the handoff happened. Of the flagged categories, distillation is the odd one out: it means extracting a model's capabilities to train a competing model, which Anthropic blocks to stop near-frontier abilities leaking out without safeguards attached. The cybersecurity classifier is the broad one. Anthropic designed it to block not just exploit development but offensive cyber tasks in general: reconnaissance, discovery, lateral movement, the agentic steps that make up a real attack. In an internal evaluation run with Fable 5 set to block rather than fall back, and which did not attempt to evade the safeguards, the classifiers stopped the model from making any progress on those tasks. One external partner found Fable 5 complied with zero harmful single-turn requests on cyberattack planning, exploit development, or defense evasion, holding up against 30 different public jailbreak techniques. The trade-off is false positives. Anthropic tuned the safeguards conservatively to ship fast, so they sometimes catch harmless requests. The company says fallback fires in under 5% of all sessions, so for more than 95%, Fable 5 behaves like the cyber-unrestricted Mythos 5. That figure covers every fallback, genuine blocks included, so it caps the total disruption rather than measuring the false-positive rate on its own. Anthropic says it will narrow the safeguards and cut false positives after launch. On robustness, the numbers are specific. An external bug bounty ran over 1,000 hours and produced no universal jailbreak, a prompt, or a harness that strips the safeguards wholesale. External red teams found none on long-form agentic tasks either, with one caveat Anthropic states plainly: the UK's AI Security Institute made progress toward a universal jailbreak within a brief initial testing window. Anthropic concedes it is likely impossible to fully prevent universal jailbreaks, and its stated goal is to make any that remain slow and costly enough to catch before they are used at scale. Why is the capability a threat The case for treating this model carefully was laid out in April, when Anthropic released Claude Mythos Preview to a limited group through Project Glasswing. The technical write-up from Anthropic's red team is the part worth reading. During testing, Mythos Preview identified and exploited zero-day vulnerabilities in every major operating system and every major web browser when a user directed it to. The oldest bug it found was a 27-year-old flaw in OpenBSD, an operating system known mainly for its security. It autonomously wrote a remote code execution exploit against FreeBSD's NFS server from a 17-year-old bug, triaged as CVE-2026-4747. Anthropic describes the result as full root for an unauthenticated attacker from anywhere on the internet; NVD's entry is more measured, noting the stack overflow itself does not require the client to authenticate, but frames kernel code execution as reachable by an attacker able to send packets to the NFS server while the kgssapi.ko module is loaded. By Anthropic's own account, it did not explicitly train these capabilities in; they emerged as a side effect of general improvements in code, reasoning, and autonomy, the same gains that make the model better at patching. The red team's flat warning: mitigations whose security value comes from friction rather than hard barriers get much weaker against a model that grinds through tedious exploitation steps at scale. Hard technical barriers like KASLR and W^X still raise the cost; the warning is narrower, aimed at defenses that lean on attacker patience or manual effort, and the model can now supply itself. Mythos 5 carries those skills forward. Anthropic says users will find it comparable to or somewhat stronger than Mythos Preview. The defender's actual problem The defensive case is not hypothetical. In the first weeks of Project Glasswing, Anthropic and roughly 50 partners used Mythos Preview to find more than ten thousand high- or critical-severity vulnerabilities in systemically important software. Cloudflare alone found 2,000 bugs, 400 of them high- or critical-severity. Mozilla found and fixed 271 in Firefox 150, more than ten times what it caught in Firefox 148 using the older Opus 4.6. Anthropic says the same pressure is visible beyond Glasswing, in vendors shipping unusually large security releases. That flood is the catch. Finding bugs is now cheap and fast. Verifying, triaging, and patching them is not, and it still runs on human time. Anthropic reports that open-source maintainers, already buried under low-quality AI-generated bug reports, have asked it to slow its disclosures because they cannot write patches fast enough. In Glasswing, it says a high- or critical-severity bug found by the model takes about two weeks to patch on average. The bottleneck has moved from discovery to the fix, and the gap between a public disclosure and a deployed patch is where attackers live. The red team's N-day experiments sharpen the point: starting from nothing but a disclosed CVE and its patch, Mythos Preview built working Linux privilege-escalation exploits in under a day each, at a few thousand dollars or less in compute. For defenders, the read is the same as ever, just on a shorter clock: assume a high-severity CVE can become a working exploit within hours of disclosure, not weeks. That means prioritizing auto-update paths for internet-facing systems and treating dependency bumps that carry CVE fixes as time-sensitive work rather than backlog. MFA and comprehensive logging stay the baseline, so a single missed patch does not become the only thing standing between an attacker and the network. Anthropic has opened a Cyber Verification Program that lets vetted security professionals use its models for legitimate offensive work without the cyber safeguards. A new 30-day data retention requirement Anthropic is also changing how it handles data for Mythos-class models. It will require 30-day retention for all traffic on Fable 5, Mythos 5, and future models at this capability level, across both first- and third-party surfaces. The company says it will not use the data for training or any non-safety purpose, will log all human access, and will delete it after 30 days except where a safety investigation or legal obligation requires holding it longer. The stated reason is defensive: the data helps detect novel attacks and jailbreaks that operate across many requests. Teams with strict data-handling requirements will want to factor that retention window in before routing sensitive traffic through these models. Anthropic plans to widen Mythos 5 access through a trusted-access program, and says that once compute capacity catches up, it aims to fold Fable 5 back into subscription plans without the usage-credit premium that kicks in after June 22. The larger question the launch raises is the one Anthropic has been circling since April: similarly capable models from other labs are coming, and not all of them will ship with a wall of classifiers in front. The defensive head start Glasswing was meant to buy only matters if the rest of the industry uses it.

Anthropic just publically released its Mythos-grade AI at no extra cost for a limited time

* Claude Fable 5 brings Mythos-grade intelligence to general users. * Safeguards reroute some risky queries to Claude Opus 4.8, triggering in * Free on Pro/Max/Team/Enterprise plans until June 23; after that it requires usage credits. It's very easy to get lost in the world of AI developments. What were once mighty AI models not intended for general use are now being rolled out for everyone to use on a daily basis. And there are no signs of development slowing down any time soon. You may have heard of Claude Mythos, a powerful LLM that wasn't meant to be used by regular users. Due to its focus on cyber security, giving everyone the keys to this mighty AI would have been detrimental. However, it was still functionality superior to other models. Now, Anthropic has revealed Claude Fable 5, a Mythos-class model designed for general use. Claude Fable 5 delivers Mythos-grade AI at no additional cost But only for a short time Over on the Anthropic blog, the company has revealed the general use of Claude Fable 5. This model is designed to be a version of its mighty Mythos model but without the same cyber security focus. That way, people can take advantage of the smartest LLM Anthropic has without giving the tools to people who will use them for evil. Of course, just because Fable 5 isn't designed for cyber security purposes, doesn't mean bad actors couldn't use it to develop malware anyway. Fortunately, Anthropic thought of this: We've therefore launched the model with safeguards that mean queries on some topics will instead receive a response from our next-most-capable model, Claude Opus 4.8. To release the model both safely and quickly, we've tuned these safeguards conservatively -- they'll sometimes catch harmless requests, though they trigger, on average, in less than 5% of sessions. Here's the good news: Anthropic has made Claude Fable 5 available for Pro, Max, Team, and seat-based Enterprise plans at no extra cost. The bad news is that it'll go away again on June 23rd and require usage credits to continue use, so get your testing in now. If things look up in the future, Anthropic says it will re-add Fable 5 to the paid plans at no additional cost. If Claude Code is going away for Pro users, I can't recommend Claude anymore This one hurts to write. Posts 42 By Mahnoor Faisal

Microsoft restricts Claude Fable for employees over data retention concerns

Anthropic released Claude Fable, its first Mythos-class AI model, yesterday and it's already causing concerns inside Microsoft. Sources tell me that Microsoft is limiting the use of Claude Fable 5 for employees because of Anthropic's new data retention requirements. While Microsoft quickly rolled out Claude Fable 5 to its GitHub Copilot and Foundry customers, I'm told the model isn't available in the model picker that Microsoft employees use for internal versions of GitHub Copilot. All other Claude models are still available internally at Microsoft, because they operate under Zero Data Retention (ZDR) rules. I understand that Microsoft has been telling employees that its legal teams are evaluating changes to Anthropic's data retention requirements. I'm told that the main concerns are around customer data and confidential information, and that it's not yet clear whether Microsoft's legal teams will clear Claude Fable for internal use. Claude Fable 5 requires data retention to operate Anthropic's new safety classifiers, meaning Anthropic retains prompts and outputs and deletes this data after 30 days. Some prompts and outputs can even be stored for up to two years if they're flagged as violating Anthropic's usage policy, so there are legal concerns around how Microsoft should be using Claude Fable 5. Claude Fable 5 is the first broad release from Anthropic's Mythos class of AI models, and it arrives just weeks after the company said the family was so capable at cybersecurity tasks that it was too dangerous to release publicly. Anthropic has put prompt safeguards in place to make Fable 5 less dangerous, leading to these challenging data retention changes. We reached out to Microsoft to comment its use of Claude Fable 5, but the company didn't respond in time for publication.

Anthropic backtracks on policy that 'sabotaged' researchers' work - Engadget

It wasn't a good look for a company that prides itself on working closely with the academic community. Anthropic is walking back a policy that discreetly hamstrung researchers using its new Claude Fable 5 LLM to create competing AI models, the company told Wired. "We're changing Fable 5's safeguards for frontier LLM development to make them visible," the company said in a statement. "We made the wrong tradeoff and we apologize for not getting the balance right." When Anthropic released Claude Fable 5, a new model based on its powerful Mythos system, researchers noted something odd. They found that that Fable 5 would quietly reroute requests to a lesser model when asked to perform certain actions. Moreover, that restriction wasn't disclosed in the model's documentation. The new model was either refusing or degrading responses for tasks like training competing LLMs, debugging AI code and optimizing neural architecture. Researchers were bothered not only by that degradation but by Anthropic's lack of transparency about it. They were also concerned, of course, that they had burned tokens and money for a model that didn't do what they expected. Anthropic has painted itself as a more ethical and researcher-friendly alternative to OpenAI, so its actions with Fable 5 created a swift backlash. "Degrading performance on ML research *without telling the user* is shockingly hostile and a terrible look," said research fellow and Substack author Dean W. Ball on X. Anthropic isn't reversing its safeguard policy on Fable 5, but rather making the restrictions visible to users. "If the company suspects a user is trying to use Claude to build a highly capable AI it will alert them that it's either refusing the request, or rerouting the user to a less capable model," Wired wrote.

Anthropic releases Mythos-like AI model to the public two months after private rollout rocked Wall Street

Two months after Anthropic rolled out Mythos to a limited number of users, citing concerns about the artificial intelligence model's potential to do damage in the wrong hands, the company said it's ready to release an equally powerful model to the public. Anthropic on Tuesday announced Claude Fable 5, a Mythos-class model that will be available to its enterprise customers and paid subscribers. The company said the broad release is possible because of new safeguards that block responses in specific high-risk areas, including cybersecurity and biology. "For us, it's really around what we call 'race to the top,' being able to provide this technology in a valuable fashion, and at the same time providing the right safety guardrails so that it can do asymmetrically more benefits than harm," Dianne Penn, Anthropic's head of product management for research, told CNBC in an interview. Anthropic captivated Wall Street and government officials in April with the unveiling of Mythos, which excels at identifying security flaws within software. The company said it did not plan to make the model generally available, and it has limited the rollout to a select group of companies as part of a cybersecurity initiative called Project Glasswing. But with the launch of Claude Fable 5, Anthropic is honoring its stated "eventual goal" to deploy Mythos-class models at scale. It's also capitalizing on growing momentum and investor interest in its technology ahead of a potentially massive IPO, which is expected to take place as soon as this year. Anthropic said Claude Fable 5 shows "exceptional performance" across software engineering and knowledge work tasks. On some benchmarks, it scored more than 10% higher than Claude Opus 4.8, another model the company announced late last month, according to a blog post. Claude Fable 5 represents a "significant jump" in capability, which is why Anthropic had to implement additional guardrails to prevent misuse, Penn said. If a user asks a high-risk question, like how to make ricin, a toxin, for instance, the model will block its response and fall back to Claude Opus 4.8 to deliver a safe answer.

Anthropic launches powerful Fable 5 model publicly, while keeping Mythos restricted over cybersecurity concerns

Serving tech enthusiasts for over 25 years. TechSpot means tech analysis and advice you can trust. In context: Anthropic's latest release is really a story about control, not just capability. The company is offering two versions of the same underlying model: Claude Mythos 5 for a small circle of trusted partners, and Claude Fable 5 for everyone else. The split reflects a core challenge Anthropic is still trying to solve - how to deploy an extremely capable system into the wild without simultaneously handing attackers a new class of offensive tools. Mythos has already shown what it can do when it is not heavily restricted. Since April, when an earlier preview was sent to about 150 organizations under the banner of Project Glasswing, users have reported more than 10,000 critical security flaws in their own systems. Those same capabilities could also be used by attackers looking to break in, rather than to patch security holes. For that reason, Mythos 5 is staying behind the glass for now. Anthropic is keeping it in the hands of a "small group of cyberdefenders and infrastructure providers," along with select biology researchers, and is coordinating with US government agencies as part of the rollout. Access is effectively on a need-to-know basis, with the company signaling that a broader "trusted access program" will come later. Fable 5 is where Anthropic is testing what a general-purpose release of Mythos-class technology looks like under constraint. Technically, it runs on the same underlying model as Mythos 5, but with hard limits built in. The system is designed to refuse or redirect a long list of requests related to cybersecurity, biology, and chemistry. When those guardrails trigger, the query is silently routed to an older model, Claude Opus 4.8, instead. Anthropic has also wired Fable 5 to watch for distillation, where a user tries to harvest large volumes of answers to train a smaller model of their own. If the system thinks that is happening, those requests are also redirected to Opus 4.8. In other words, the company is not only trying to control what the model will talk about, but also what others can learn from it. Anthropic has been wrestling with these decisions for months. Diane Penn, the company's head of product management, told Wired that testing and feedback since the April preview have helped shape the current strategy, even though it is still far from perfect. "We're trying to make improvements in a way that's beneficial, even if we don't have the perfect [solution] for every use case to start," she says. "Out of all the different approaches, this emerged as the most viable and the best one. We just ended up feeling like this was the best product choice for users to get the maximum value out of Fable 5." For now, the filters are tuned to err on the side of over-blocking. Penn has acknowledged that some harmless queries will be routed to the older model. Anthropic says it wants to refine its classifiers over time but argues that this level of caution is the only way to justify a wider release at this stage. The stakes are higher because Fable and Mythos are not just chatbots that respond to prompts and stop. Anthropic says both can run "unattended" for longer stretches than previous Claude models, carrying out sequences of instructions without constant supervision. That shift toward more agent-like behavior could substantially boost software engineering and other technical work, especially given Fable 5's stronger code generation and visual capabilities. But it also raises obvious questions about what happens if those capabilities are misused. Anthropic's pricing reflects how powerful it believes these systems are compared with its other models. Fable 5 and Mythos 5 cost $10 per million input tokens and $50 per million output tokens, roughly double the company's other public models but still cheaper than the earlier Mythos Preview. The higher price reflects both the performance gains and the sense that these models are still positioned as specialized systems, not yet just another SKU in a growing catalog. Around Anthropic, competitors are moving in a similar direction. OpenAI has rolled out its own advanced cybersecurity model to a small circle of partners and convened a working group that echoes Project Glasswing. Both companies are preparing for potential IPOs and are under pressure to show investors they can ship cutting-edge technology without triggering backlash over safety concerns. Even some of the people watching from the outside say the unease is justified. Canadian finance minister François-Philippe Champagne told the BBC that public concern around Mythos stemmed from "it's the unknown, unknown." Anthropic co-founder Jack Clark has made a similar point from the inside, arguing that the industry has not yet figured out how to slow itself down. "You want the option to be able to take your foot off the gas and put your foot on the brake," he said. "Right now, it's like the AI industry has a gas pedal, but it doesn't have a brake pedal."

Anthropic Releases 'Safe' Version of Its Mythos A.I. Technology

Called Claude Fable 5, it is twice as expensive as the company's previous flagship system. In April, Anthropic unveiled a new artificial intelligence system called Claude Mythos. But the company said it couldn't share its new A.I. with the public because Mythos could become a powerful tool for hackers looking for ways to break into computer networks. That warning caused panic among executives and government officials who worried that the technology was about to usher in a new era of cybersecurity threats. Now Anthropic is releasing a straitjacketed version of Mythos that researchers believe is safe for widespread use and relies on older technology for some of what it does, the company said on Tuesday. Called Claude Fable 5, this new system includes additional guardrails designed to block responses related to cybersecurity, biology and other vulnerable areas. Because of these guardrails, hackers may struggle to attack computer networks using Fable. But businesses and cybersecurity experts may also struggle to defend networks using the new system. Most queries from Claude users in areas that could be perceived as too risky, the company said, will be handled by Claude Opus 4.8, which was released last month and was also designed to avoid the security risks of Mythos. "That is the core of why we have been able to release Fable," said Anthropic's head of product management, Dianne Penn. Anthropic shared the initial version of Mythos with about 40 organizations that maintain critical computer infrastructure so they could use the system to patch security vulnerabilities before hackers exploited them. This month, Anthropic said it would share Mythos with 150 additional organizations. Cybersecurity experts disagreed on whether Anthropic was right to limit the release of the technology. Some applauded the company for restricting access to Mythos, while others criticized Anthropic for not sharing it with a wider pool of researchers who could try it, get a handle on what it can and cannot do and start using it for defense. Mythos, which is still available to a limited number of organizations, does not have the guardrails built into Fable. Outside researchers have shown that guardrails placed on A.I. systems are not always reliable, a point that Anthropic researchers recognize and say they are working on. "This is why we have invested so much in safety research," Ms. Penn said. "It is a continuing progression." Fable tops all other publicly available A.I. technologies in overall performance, according to benchmark tests from Vals AI, a company that tracks the performance of the latest A.I. technologies. Its overall performance is 5 percent higher than Claude Opus 4.8, according to Vals AI's tests. The new system is also twice as expensive as Opus 4.8. Fable is particularly adept at generating computer code and in mathematics, said Rayan Krishnan, the chief executive of Vals AI. But other systems still outperform Fable in other areas, including tasks involving health care and tax evaluation. Over the past several months, companies in the United States, China and other parts of the world have built A.I. systems that are particularly good at writing computer code. If an A.I. system can write code, it can potentially identify vulnerabilities in software applications. Hackers can then use the technology to attack computer networks. But tech companies, other businesses and cybersecurity experts can also use the technology to defend networks. A week after Anthropic unveiled Mythos, the company's chief rival, OpenAI, said it was sharing similar technology with a limited group of partners. But the company shared its model, GPT-5.4-Cyber, with a much larger group -- hundreds of organizations. It said it would then release it to thousands more partners in the coming weeks. (The New York Times sued OpenAI and Microsoft in 2023, claiming copyright infringement of news content related to A.I. systems. The two companies have denied those claims.)

Claude Mythos: Anthropic releases version of AI tool despite risk concerns

A version of an artificial intelligence (AI) tool which the company said was too powerful to be released to the public has just been released to the public. Claude Fable 5 is a version of Anthropic's Claude Mythos, an AI program which caused a stir among technology, finance, and government leaders when it was released privately in April for previewing and testing. Some senior figures worried the tool was so powerful it could pose financial security risks, though others have questioned how much of the hype is marketing spin. Anthropic said on Tuesday Fable will be released with safeguards and user limitations in place, though it said "releasing a model this capable comes with risks". "Fable's capabilities exceed those of any model we've ever made generally available", it added. When Mythos was first released to a small group of organisations to preview the tool, Anthropic said it was doing so because the tool was so intelligent that it could be dangerous. Anthropic is expected to become a public company soon, as its private valuation has neared $1tn (£747bn). The company also said on Tuesday said that the roughly 150 groups that had been given access to preview Mythos will now have access to Claude Mythos 5, which does not have limitations on cybersecurity or biology, depending on an organization's specific uses. Anthropic said this access was limited to a "small group of cyberdefenders and infrastructure providers" but is expected to go beyond that soon. "We intend to expand access to Mythos 5 through a broader trusted access program", the company said. Both Fable and Mythos, which are essentially the same model but with different safeguards and levels of access in place, can work "unattended" on human commands that the tools are given for longer periods of time "than any previous Claude models." Anthropic added. Anthropic co-founder Jack Clark last week told BBC Newsnight that the ability of AI tools was expanding so rapidly that the company thought there should be a way for the public to slow the technology's advancement. "You want the option to be able to take your foot off the gas and put your foot on the brake", Clark said. "Right now, it's like the AI industry has a gas pedal, but it doesn't have a brake pedal."

Anthropic spins a Fable of a neutered, safer Mythos

It's here. Anthropic's Mythos model, supposedly too dangerous for public release in April, is now available to wreak havoc or tackle other tasks for a hefty price and with some new guardrails in place. Just make sure you don't mind having Anthropic keep some of your data for a while. The AI biz on Tuesday announced public availability of Claude Fable 5, a Mythos-class model, and private availability of Claude Mythos 5 for Glasswing partners. Both are distinct from Mythos Preview, the model family's elder sibling. Mythos-class models are said to be a tier above Claude Opus in terms of benchmark performance. "Fable 5's capabilities exceed those of any model we've ever made generally available," the company said. "It is state-of-the-art on nearly all tested benchmarks of AI capability, showing exceptional performance in software engineering, knowledge work, vision, scientific research, and many other areas." As if to suggest that the current technocracy still has room for humanities scholars, Anthropic notes that the name Fable comes from the Latin fabula, "that which is told," similar to the Greek mythos. "The safeguards are what distinguish the two models (Fable and Mythos) and are why we've given them different names," the company said. Anthropic considers Fable capable of causing serious damage without safeguards. So part of its protocol is to failover to Opus 4.8 for certain types of queries - prompts related to cybersecurity, biology and chemistry, or distillation. Fable 5 ships with a new set of classifiers, which are separate AI models that look for misuse. The AI biz is also instituting a new data retention policy - it's retaining log data to have a record in the event of misuse. This applies specifically to organizations with zero data retention policies - which now technically are not zero data retention. "Prompts submitted to, and outputs generated by, Mythos-class models are retained for 30 days for trust and safety purposes, on every platform where these models are offered," the company explains on its updated support page. Consumer plans (Claude Free, Pro, and Max) are unaffected - because Anthropic already has a data retention policy in place for those tiers. "This change only applies to organizations that have set up workspaces with zero data retention (ZDR) in Claude Console, use Claude Code with ZDR in Claude Enterprise, or access Claude through AWS Bedrock, Google Cloud Agent Platform, or Microsoft Foundry with ZDR," Anthropic's documentation explains. Temporarily retaining prompts and outputs is necessary to prevent misuse, Anthropic insists, adding that it won't use this data for training new models. In terms of benchmarks, the two models significantly outperform Anthropic's own Opus 4.8, Anthropic's Codex 5.5, and Google's Gemini 3.1 Pro on the chosen test suite. Not listed is Google's more recent Gemini 3.5 Flash [PDF]. Fable 5 is said to excel at software engineering. Anthropic claims that Stripe used the model to migrate its 50-million line Ruby codebase in just one day. That project, Anthropic claims, would have taken a Stripe team two months without AI help. Fable 5 is also supposedly more token efficient than than its predecessors, which might somewhat mitigate the higher token cost. Fable 5 and Mythos 5 both charge $10 per million input tokens and $50 per million output tokens, which is less than half the price of Claude Mythos Preview ($25/m input, $125/m output). Mythos 5 delivers the strongest results yet for resisting prompt injection attacks, which saw a success rate of 4.8 percent over 100 attempts, per the model safety card [PDF]. That's comparable to Claude Opus 4.8. As of today, until June 22, Fable 5 is available through Pro, Max, Team, and seat-based Enterprise plans at no extra cost. After that, Fable 5 will no longer be available through those plans and will require usage credits. At some point thereafter, "if capacity allows," Anthropic intends to restore Fable 5 access in subscription plans. ®

Claude Fable 5 curbs: aimed at China, hit AI researchers

The restrictions built to keep China's AI labs out of Anthropic's most powerful model held firm. The way it first enforced them lasted barely two days. Anthropic built the Claude Fable 5 curbs to keep China's AI labs out of its most powerful public model. The loudest complaints came from its own side of the firewall. On 9 June, Anthropic released Claude Fable 5, a tamed, public-facing version of Mythos, the model it had withheld since April over its talent for finding and exploiting security flaws. Fable 5, the company said, was "safe for general use." Built into it was a set of classifiers that flag requests touching cybersecurity, biology, chemistry, distillation and frontier model development, then quietly route them to Anthropic's second-best model, Claude Opus 4.8. The word that caused the trouble was "quietly." According to Anthropic's own system card, the interventions limiting "frontier LLM development" would "not be visible to the user." AI researchers said they were being silently downgraded, or shadowbanned, the moment Fable 5 decided their machine-learning work looked too much like an attempt to build a rival. "It felt like Anthropic was saying to the public, we don't trust anybody else to do AI research," Will Brown, research lead at Prime Intellect, told Wired. The research firm SemiAnalysis said the model was degrading its GPU-inference work. The objections came from open-source advocates who usually call Anthropic too closed, and from safety researchers who usually defend it. Anthropic reversed within two days. "We're changing Fable 5's safeguards for frontier LLM development to make them visible," it said. "We made the wrong trade-off and we apologize for not getting the balance right." The fix was disclosure, though. The restriction stayed. What the Claude Fable 5 curbs still do to China On access, the experts think the restriction works. "Chinese AI developers might find it nearly impossible now to use Anthropic's latest model to accelerate their own model development," Kyle Chan, a fellow at the Brookings Institution, told the South China Morning Post. Claude has never been officially sold in China, but developers there long relied on workarounds. Fable 5's classifiers are much harder to route around. Anthropic says those workarounds were industrial. In February it accused DeepSeek, Moonshot and MiniMax of running more than 16 million exchanges through roughly 24,000 fraudulent accounts to distil Claude's coding and reasoning. The White House took up the theme in April, calling the campaigns industrial-scale theft. The dependency runs both ways Then the picture inverts. US firms are increasingly reaching for China's DeepSeek because it is cheap, while Chinese developers keep buying grey-market Claude through proxy "transfer stations" that resell it on Taobao and Telegram at as little as a tenth of list price, according to the Oxford China Policy Lab. Beijing is busy building walls of its own: this week Meta began dismantling its $2 billion purchase of Manus, the Chinese-founded, Singapore-based startup, after regulators ordered the closed deal reversed. Critics see less principle than protectionism, noting that Anthropic trained Claude on scraped, copyright-protected text before it began policing anyone else's extraction. The curbs also land as Anthropic heads for the public market, having filed confidentially for an IPO at a $965bn valuation. A model too open invites distillation; a model too closed invites the backlash Fable 5 just triggered. Anthropic has drawn its line. Whether classifiers can hold it, when the first version lasted barely two days, is the question its rivals on both sides of the Pacific are already testing.

Anthropic Apologizes For One of the Guardrails on Its Fable 5 Model, and Will Change It

Anthropic's Fable 5 model is the nerfed version of Mythos, which is in turn the model so scarily powerful that it could ostensibly endanger the world if it were released without guardrails. Most of the guardrails, especially the ones designed to prevent users from using Fable to build cyber- or bio-weapons, are very noticeable. But one guardrail, aimed at preventing users from using Fable 5 to train other AI models, was invisible, which sparked unusual displays of user outrage. And now Anthropic has asked for take-backs. The controversial invisible guardrail will be made visible. In a statement to Wired, Anthropic wrote "We're changing Fable 5's safeguards for frontier LLM development to make them visible." "We made the wrong tradeoff and we apologize for not getting the balance right," the statement added. In the model's system card, Anthropic was upfront about what it was trying to do: "Unlike our interventions for cybersecurity, biology and chemistry, and distillation attempts, these safeguards will not be visible to the user. Fable 5 will not fall back to a different model. Instead, the safeguards will limit effectiveness through methods such as prompt modification, steering vectors, or parameter-efficient fine-tuning (PEFT)." In other words, when Fable 5 prompts showed the telltale signs of a user developing a frontier LLM, instead of doing what it does with prompts about biology, chemistry, or cybersecurity and switching to an inferior model, or simply refusing the request, it was silently changing the prompt in order to generate faulty results with the potential to hamper the user's model development. Using the model to train another model is against Anthropic's terms of service, but users still felt like this measure was a violation of users' trust. Reddit user CheatCodesOf Life put it this way: "I wouldn't use this thing for anything to be honest. A refusal or HTTP-4xx error for content is fair enough, but this is basically taking your money and poisoning your code base."

Anthropic's too-scary-to-release AI hacking tool is actually coming out -- kind of

Both models show advanced capabilities for tackling even highly complex analytical tasks. By now, you've almost certainly seen at least one AI demo that left you feeling a little off-kilter -- "That's not just good; that's scarily good." Whether we're talking about AI stealing jobs from artists, or generating novel proofs for mathematical mysteries, we're constantly finding new ways where AI-powered systems deliver results that make them seem almost uncomfortably powerful. Recently, we've heard Anthropic sharing its concerns about releasing its latest Claude Mythos model, which is just a little too good at finding software vulnerabilities. Today, however, Anthropic starts moving forward with just that. Automated bug-hunting is nothing new, and computer scientists have been using tools like fuzzers for years, flooding software with random inputs in the hopes of causing a glitch. But AI represents a much more potent threat, and as vulnerability finders like Mythos have evolved, their creators have grown appropriately apprehensive about sharing their most powerful models. Rather than let its tech sit around gathering dust forever, today Anthropic shares the compromise it's sorted out for how to deploy Mythos responsibly and safely. The key to this approach is dividing things into two separate models: Claude Fable 5 and Claude Mythos 5. Fable 5 is the model for the general public, and it's a full-featured solution for tackling analytical tasks. This doesn't have to be just finding software vulnerabilities -- Fable 5 can help code, has powerful vision analysis tools, and can even develop internal strategies over time. According to Anthropic, "Fable 5's capabilities exceed those of any model we've ever made generally available." But Anthropic is also releasing it with a few strings attached. Fable 5 is designed to resist attempts to use it to find the latest zero-day exploit, ready to be taken advantage of by bad actors and wreak havoc on global computer systems. When someone tries to push Fable 5 outside those bounds, the model will fall back to Claude Opus 4.8, instead. But then there's Mythos 5, which is internally the same as Fable 5, but lacks many of those extra safeguards. The key here is that Anthropic plans to release it highly selectively, only inviting trusted members of the cybersecurity community to use it. In order to stay ahead of the bad guys, it's important for a tool like this to be used to identify bugs and fix them, and this solution is intended to mitigate the risks that Mythos 5 gets re-purposed for nastiness. Anthropic has developed a series of "classifiers" that attempt to recognize what users are trying to do with Fable 5, and they don't just try to block hacking. They're also built to stop anyone from using the model to develop ways to synthesize dangerous chemicals or biological compounds, or to extract enough internal information to build their own Fable 5 -- without the safeguards in place. Hopefully that proves enough to keep this public Mythos-based release operating safely, because we're sure a lot of people will be trying very hard to unlock its full capabilities.

Anthropic finally releases Mythos to the public, but it's so heavily guarded it barely works

Anthropic has been touting the power of its Claude Mythos AI model to find and help fix security vulnerabilities, and it has finally released a public version, Fable 5. However, some cybersecurity experts are complaining that the guardrails meant to prevent abuse are stifling its real-world usefulness. Fable is billed as a "Mythos-class" model that can conduct reasoning, coding, and cybersecurity tasks better than both the Mythos preview release and conventional AI models, including OpenAI's GPT 5.5 and Google's Gemini 3.1 Pro. This ideally helps security teams pinpoint flaws, study issues, and develop fixes with relative speed. Claude Price $20 See at Claude Expand Collapse The differences between Fable 5 and its Mythos 5 equivalent come down to the "safeguards," according to Anthropic. If the Fable model believes there's a risk of abuse, such as writing malware, it reverts to Claude Opus 4.8. Mythos 5 doesn't have as many protections, and access is limited to members of the Project Glasswing initiative meant to secure vital software with the help of the U.S. government and certain companies, including Apple, Google, Microsoft, and hundreds of others. The new Fable and Mythos versions are available at respective prices of $10 and $50 per million input tokens, or less than half the price of the Mythos preview. Pros who want fewer limits on any of Anthropic's models have to get approval through a Cyber Verification Program. Why is Fable 5 so bad for some cybersecurity experts? They feel it's simply too safe Anthropic is clear that it set Fable's guardrails "conservatively" for the sakes of both a quick launch and safety, and that there will be moments where "harmless requests" limit functionality. It expects to improve these protections and limit mistakes going forward. As TechCrunch explains, however, cybersecurity veterans still aren't thrilled. Experts at IBM and elsewhere have pointed out that asking to write or review secure code downgrades to Opus, and that the limits might kick in simply by using the 'wrong' keyword. For some, it could be difficult to use Fable the way it was intended. 5 AI Technologies Hackers Can Use in Terrible New Ways Some people think The Matrix is an instruction manual. Posts By Sydney Butler The initial caution might be warranted. AI can discover and use vulnerabilities more easily and quickly than before, and that could lead to zero-day exploits that developers haven't had time to fix or even analyze. Deals Discounted AI software deals for security and dev teams Explore savings on AI and cybersecurity software deals, discounts on model access, subscription plans, cloud credits, code-scanning tools, and security training bundles. Shop Software, AI & Subscriptions offers to outfit dev and security teams affordably. Deals Explore Software, AI & Subscriptions Deals Microsoft, for instance, marked a record "Patch Tuesday" in June where it fixed almost 200 security bugs as its team and the community used AI to catch issues they'd otherwise miss. If attackers could use tools like Fable or Mythos without limits, there's a risk they could routinely compromise systems before patches are ready, and with relatively little effort.

Claude Fable 5 brings Mythos-class capabilities to public users

Anthropic has opened the doors to its most advanced AI system yet, but not without placing guardrails around its most sensitive capabilities. The company on Tuesday unveiled Claude Fable 5, the first publicly accessible version of its Mythos-class AI model. Anthropic says the model outperforms every generally available system it has released to date, delivering major gains in software development, analytical work, visual understanding, and scientific reasoning. The launch marks a significant shift for the AI firm. Until now, Mythos-level capabilities remained restricted to a small group of trusted organizations because of concerns that the technology could aid cyberattacks and other harmful activities. Fable 5 changes that equation by bringing those capabilities to a broader audience while routing certain high-risk requests through a less permissive model. Anthropic said Fable 5 includes a new layer of AI-powered classifiers designed to detect potentially dangerous prompts and jailbreak attempts. The company stress-tested those safeguards before releasing the model to the public. If users submit requests involving cybersecurity, biology, chemistry, or model distillation, the system automatically redirects those interactions to Claude Opus 4.8 instead of Fable 5. Users receive a notification whenever that fallback occurs, allowing them to understand why the switch happened. The company acknowledged that the safeguards may occasionally block harmless requests. However, Anthropic said early testing showed that fewer than 5% of sessions triggered a fallback response. "We've deliberately tuned the safeguards to be cautious," the company said, adding that it plans to refine the system over time to reduce unnecessary interventions. Anthropic argues that Fable 5 delivers its biggest advantages when tackling lengthy and complex assignments. The company says the model can stay focused across extended workflows while maintaining consistency over large amounts of context. During early evaluations, payments company Stripe reported that the model completed a large-scale migration across a 50-million-line Ruby codebase in roughly one day. According to Anthropic, the same task would have otherwise occupied an engineering team for more than two months. The company also highlighted gains in financial analysis. On Hebbia's finance benchmark, Fable 5 posted the strongest results among leading models. Trading firm IMC found that the system performed well across factual analysis, reasoning tasks, root-cause investigations, and expected-value calculations. Fable 5 also advanced Anthropic's vision capabilities. The model can extract detailed information from scientific figures and reconstruct web applications from screenshots. In another demonstration, it completed Pokémon FireRed using a minimal vision-only setup after earlier Claude models struggled with the challenge. Alongside Fable 5, Anthropic introduced Claude Mythos 5, a version of the same underlying model with fewer restrictions in select areas. Anthropic says the model offers the strongest cybersecurity capabilities it has developed so far. Access to Mythos 5 remains limited. The company is deploying it through Project Glasswing while expanding availability to approved cyber defenders and organizations responsible for critical infrastructure. Anthropic said it intends to broaden access further through a trusted program in the future. Fable 5 is available through Anthropic's Claude API and consumption-based Enterprise offerings. Through June 22, Pro, Max, Team, and seat-based Enterprise subscribers can access the model at no additional cost. After that date, Anthropic plans to move users to a credit-based system before eventually restoring broader subscription access. As competition intensifies across the AI industry, Anthropic is betting that powerful models do not have to arrive without safeguards. With Fable 5, the company is testing whether frontier capabilities and tighter controls can coexist in the hands of everyday users.

Anthropic releases its first Mythos-class model Claude Fable

Anthropic just announced Claude Fable 5, a new AI model it said is the most powerful model it has ever made widely available. According to the company, Fable 5 "shows exceptional performance in software engineering, knowledge work, and vision," with its lead over other models growing as tasks become longer and more complex. Fable 5 marks the first broad release from Anthropic's Mythos class of AI models, after the company said the family was so capable at cybersecurity tasks that it was too dangerous to release publicly. Anthropic said the release was "made possible by new safeguards that block responses in specific high-risk areas," with the system falling back to Claude Opus 4.8 -- a model it praised for "honesty" when it launched last month. Anthropic singled out cybersecurity and biology as two domains where the safeguards may block responses, both areas widely considered sensitive topics for advanced AI systems. The company said that in testing, 95 percent of Fable sessions ran entirely on Fable responses, without falling back to Opus 4.8. The company is also releasing Claude Mythos 5, but provided few details on what that means. In a blog, Anthropic said Mythos 5 is the same underlying model as Fable 5, "but with the safeguards lifted in some areas." For now, access appears limited to the steadily expanding group of organizations granted access to Claude Mythos Preview through Anthropic's -- not entirely watertight -- private Project Glasswing initiative. Those users will be able to upgrade to Mythos 5, Anthropic said, adding that it plans to "expand access over time through a more systematic trusted-access program." Anthropic did not respond on the record to The Verge's request for comment explaining how either model relates to Claude Mythos Preview or why the models are numbered "5" when there do not appear to be any previously released Mythos or Fable models. Pricing for both models is significantly higher than its former flagship model -- double rates for Claude Opus 4.8, though it's half what users pay for Mythos Preview -- at $10 per million input tokens and $50 per million output tokens, Anthropic said.

Anthropic's Fable AI brings the capabilities of its unreleased Mythos model to regular users - Engadget

Claude subscribers can try the model until June 22 without spending usage credits. Anthropic has just announced Fable, the start of a new family of models that brings many of the capabilities of its Mythos system to the public. As a refresher, Mythos is the state-of-the-art model Anthropic debuted at the start of April through Project Glasswing. The project saw Anthropic share access to the model with select partners, including Apple and NVIDIA, with the aim of helping those organizations harden their software against AI cyberattacks. Glasswing also prompted the White House to rethink its policy on AI regulation. Last week, Anthropic said it was working as quickly as possible to release a Mythos-capable model with "highly robust safeguards" to the public, and today it's doing exactly that with Fable. Anthropic says the new system, which starts at version five to align with the company's other models, offers capabilities that "exceed" those of any it has made publicly available in the past. "The longer and more complex the task, the larger Fable 5's lead over our other models," Anthropic says. In the company's own testing, Fable beat not only its previous flagship offering, Opus 4.8, but also GPT-5.5 and Gemini 3.1 Pro from rivals OpenAI and Google. Anthropic adds Fable is particularly adept at tasks like software engineering, drug design and document analysis, but also excels in domains where the company's past models have typically underperformed. Vision is one such area. You may recall last year Anthropic streamed Claude trying (and failing) to beat Pokémon Red. The model Anthropic used for that Twitch series, 3.7 Sonnet, required an overlay to allow Claude to make sense of the game's pixel art interface and world. If you watched the stream closely, you would have noticed each time Claude moved the game's protagonist it had to reevaluate their position afterward. Between frames, Claude was essentially blind, and it couldn't "hear" when the player character hit an obstacle like a tree. By contrast, Fable can beat FireRed, the 2004 Game Boy Advance remake, with a "minimal, vision-only harness." In practice, the company says Fable's new capabilities translate to a model that can "extract precise numbers from detailed scientific figures and can perform complex vision-based tasks like rebuilding a web app's source code from screenshots alone." As for those safeguards I mentioned earlier, Claude will automatically route prompts on some topics through the less-capable Opus 4.8. "To release [Fable 5] both safely and quickly, we've tuned these safeguards conservatively -- they'll sometimes catch harmless requests, though they trigger, on average, in less than 5 percent of sessions," the company said. "With more capable models arriving in the coming months, we're working to improve our safeguards and reduce false positives as quickly as we can." On Tuesday, Anthropic also released Mythos 5, which runs on the same underlying model as Fable 5 but comes with fewer safeguards. The company will initially offer Mythos 5 through Project Glasswing before making it more widely available through a trusted access program. Claude subscribers can try Fable 5 until June 22. Thereafter, using Fable 5 will cost usage credits, which Anthropic has priced at $10 per million input tokens and $50 per million output tokens. "After this point -- when sufficient capacity allows us to do so -- we aim to restore Fable 5 as a standard part of subscription plans. We intend to do this as quickly as we can."

Anthropic just released public Mythos-class AI model called Claude Fable, details here

Back in April, Anthropic unveiled its Claude Mythos AI model that it said was too powerful to publicly release. Instead, the company has shared access with software vendors, including Apple, in an effort to use the model for enhanced cybersecurity. Now, as promised, Anthropic has released what it calls a Mythos-class model that customers can actually use. The new model is called Claude Fable. Claude Fable 5 and Claude Mythos 5 Anthropic calls Claude Fable 5 its first Mythos-class model that's publicly available for customers. "Fable 5's capabilities exceed those of any model we've ever made generally available," the company says. "It is state-of-the-art on nearly all tested benchmarks of AI capability, showing exceptional performance in software engineering, knowledge work, vision, scientific research, and many other areas. The longer and more complex the task, the larger Fable 5's lead over our other models." Anthropic adds that its new Claude Fable 5 model comes with specific safeguards, given its capabilities: Releasing a model this capable comes with risks. Without safeguards, Fable 5's capabilities in areas like cybersecurity could be misused to cause serious damage. We've therefore launched the model with safeguards that mean queries on some topics will instead receive a response from our next-most-capable model, Claude Opus 4.8. To release the model both safely and quickly, we've tuned these safeguards conservatively -- they'll sometimes catch harmless requests, though they trigger, on average, in less than 5% of sessions. With more capable models arriving in the coming months, we're working to improve our safeguards and reduce false positives as quickly as we can. Beyond Claude Fable 5, Anthropic also details Claude Mythos 5, which is basically Fable 5 without a lot of the safeguards. Mythos 5 is only available to a "small group of cyberdefenders and infrastructure providers," per Anthropic. The company says both Fable 5 and Mythos 5 cost less than half Claude Mythos Preview's price at $10 per million input tokens and $50 per million output tokens. The new models perform favorably compared to Claude Opus 4.8 and GPT-5.5, according to Anthropic's results. Claude Fable 5 arrives for customers starting today. Anthropic says it expects Fable 5 demand to be "very high, and difficult to predict." For that reason, subscription plans will be able to access Fable 5 in stages, per Anthropic * From today through June 22, Fable 5 is included on Pro, Max, Team, and seat-based Enterprise plans at no extra cost. * On June 23, we'll remove Fable 5 from those plans. Using it after that will require usage credits. If capacity allows, we'll extend the included window. * After this point -- when sufficient capacity allows us to do so -- we aim to restore Fable 5 as a standard part of subscription plans. We intend to do this as quickly as we can. You can learn much more about Anthropic's new Claude Fable 5 model here.

Claude Mythos goes public in new Fable 5 model that's 'safe for general use'

As AI models get more powerful, there's also going to be a level of danger. Anthropic announced Mythos not long ago as a model that was too dangerous for public availability, but a new version of that, Claude Fable 5, is now available and, apparently, "safe for general use." In a post today, Anthropic explains that Claude Fable 5 is a "Mythos-class" model that's available for public use and exceeds the capabilities of any other generally available AI model Anthropic has ever published. Fable 5's capabilities exceed those of any model we've ever made generally available. It is state-of-the-art on nearly all tested benchmarks of AI capability, showing exceptional performance in software engineering, knowledge work, vision, scientific research, and many other areas. The longer and more complex the task, the larger Fable 5's lead over our other models. The fear with Mythos was mainly around cybersecurity, as the model was quickly found to be able to find and use exploits in the digital world. With Fable, Anthropic says that safeguards are in place and that, if a topic comes up that Fable isn't allowed to interact with, Claude Opus 4.8 will step into the conversation instead. Anthropic says it has tuned this "conservatively," so it might trigger more often than you'd expect. Meanwhile, Claude Mythos 5 will be available to "a small group of cyberdefenders and infrastructure providers" as the same model as Fable 5, but with "safeguards lifted in some areas." Mythos 5 will initially be deployed through Project Glasswing, in collaboration with the US government, as an upgrade to Claude Mythos Preview. It has the strongest cybersecurity capabilities of any model in the world. Soon, we intend to expand access to Mythos 5 through a broader trusted access program. Anthropic adds that the "capabilities of models like Fable 5 and Mythos 5 have the potential to do profound good for the world," and shows some drastic gains in tasks like agentic coding, tool use, and cybersecurity compared to the Mythos Preview, Opus 4.8, GPT 5.5, and Gemini 3.1 Pro. The release of Fable 5, though, will be a bit tricky. Anthropic says that, as of today, Fable 5 is available to all Pro, Max, Team, and seat-based Enterprise plans, but only through June 23. At that point, Fable 5 will shift to a usage credit-based plan with the hope of restoring standard access "when sufficient capacity allows." * From today through June 22, Fable 5 is included on Pro, Max, Team, and seat-based Enterprise plans at no extra cost. * On June 23, we'll remove Fable 5 from those plans. Using it after that will require usage credits. If capacity allows, we'll extend the included window. * After this point -- when sufficient capacity allows us to do so -- we aim to restore Fable 5 as a standard part of subscription plans. We intend to do this as quickly as we can.

Claude's 'too dangerous' AI model is finally public. But there's a catch

The release includes conservative safeguards due to extreme cybersecurity capabilities that could potentially be misused, with a less-restricted version available only to select cyberdefenders. Yes, it's finally happened. A "Mythos-class" Claude model has arrived for the general public. But it's not the same "too-dangerous-for-you" Mythos you've been hearing about, and Claude subscription users will only get a limited crack at it. Claude Fable 5 is now available in the Claude desktop app, and the new model is so powerful that it has "the potential to do profound good for the world," Anthropic said in a press release. At the same time, Fable 5's extreme cybersecurity capabilities "could be misused to cause serious damage," Anthropic warns. That's why the model arrives with conservative safeguards that may "sometimes catch harmless requests." (Anthropic pegs the frequency of over-cautious denials at less than 5 percent of sessions.) Anthropic is also releasing Claude Mythos 5, based on the same underlying model as Fable 5 but with safeguards lifted in some areas, for a "small group of cyberdefenders and infrastructure providers." Anthropic has released benchmarks for Fable 5 that promise substantial improvements in coding, spatial reasoning, and other areas over previous models. But the biggest leaps come in cybersecurity, which was what all the fuss was about when news of Claude Mythos first broke. There's good news and bad news when it comes to Fable 5 access for paid Claude users. The good news is that those of us who are subscribers (versus API users who pay by the token) can use Fable 5 right now, and that includes Claude Pro, Max, Team, and seat-based Enterprise users. The bad news? Anthropic will yank Fable 5 from Claude plans on June 23rd. After that, you'll need paid usage credits to access Fable 5. (Credits are pre-paid and charged at the same rate as API usage.) Anthropic says it "aim[s] to restore Fable 5 as a standard part of subscription plans ... as soon as possible" but only "when sufficient capacity allows us to do so." In the meantime, those on paid Claude plans can kick the tires on Fable 5, but expect your usage meters to drain dramatically faster than they do with Claude Opus 4.8, the next-most-powerful Claude model. For my part, I tried using Fable 5 with Claude Code but was almost immediately kicked off, with a vague error message saying there was an "issue with this model." Hmm.

Claude Fable 5 is here -- and it's based on a model Anthropic once deemed too risky for the public

This version comes with guardrails that block some of its most powerful capabilities Back in April, Anthropic surprised the AI world when it revealed Claude Mythos, a model so capable at cybersecurity tasks that the company decided not to release it publicly. Instead, Anthropic limited access to a small group of organizations through Project Glasswing, a program created in partnership with the U.S. government and focused on cyberdefense and critical infrastructure protection. At the time, the message was clear: some AI capabilities were simply too powerful to put into everyone's hands. Now Anthropic is changing course -- sort of. The company has launched Claude Fable 5, a new model built on the same underlying system as Mythos 5. But unlike Mythos, Fable 5 is available to the public. The catch is that Anthropic has wrapped it in a series of safeguards designed to prevent users from accessing some of the model's most powerful capabilities. In many ways, Claude Fable 5 may be the most important AI release of the year because it signals a new strategy for deploying frontier models. Making Claude safer for the general public to use According to Anthropic, Claude Fable 5 and Claude Mythos 5 are fundamentally the same model underneath. The difference is that Fable includes guardrails that intervene when users enter sensitive territory such as cybersecurity, biology, chemistry and certain forms of model extraction. When that happens, requests are redirected to a less capable model instead. Anthropic says the system works in at least 95% of sessions without needing to fall back to another model, suggesting most users will never notice the restrictions. The result is effectively a consumer version and a restricted version of the same frontier AI system. That's new. Historically, AI companies have released a single model and adjusted safety filters around it. Anthropic is now drawing a more explicit line between what the public gets and what trusted organizations receive. The takeaway What's most interesting isn't Claude Fable 5, but what comes next. For years, AI companies have debated whether increasingly capable models should be released openly, restricted entirely or carefully controlled. Anthropic may have just introduced a fourth option: release the model publicly, but selectively disable the capabilities considered most dangerous. If that approach works, we may see more frontier AI systems arrive in two forms -- a public version for everyday users and a restricted version reserved for governments, researchers and vetted organizations. Follow Tom's Guide on Google News and add us as a preferred source to get our up-to-date news, analysis, and reviews in your feeds. Subscribe to Tom's Guide on YouTube and follow us on TikTok.

Anthropic releases Claude Fable 5, its first Mythos-class model for the public

Anthropic launched Claude Fable 5, its first publicly available Mythos-class model, with safeguards that block high-risk queries and fall back to Opus 4.8. Anthropic on Tuesday released Claude Fable 5, a model built on the same architecture as its restricted Mythos system, making Mythos-class intelligence publicly available for the first time. Fable 5 is available to enterprise customers and paid subscribers, but it comes with new safeguards that block responses in cybersecurity, biology, chemistry, and what Anthropic calls "distillation" scenarios, falling back to Claude Opus 4.8 to deliver a safe answer instead. Early data from Anthropic indicates the fallback triggers in fewer than 5% of sessions, meaning most users will interact with the full Fable 5 model for the vast majority of their queries. "For us, it's really around what we call 'race to the top,' being able to provide this technology in a valuable fashion, and at the same time providing the right safety guardrails so that it can do asymmetrically more benefits than harm," Dianne Penn, Anthropic's head of product management for research, told CNBC. The model represents what Penn described as a "significant jump" in capability over Opus 4.8, which is why the additional guardrails were necessary. On some benchmarks, Fable 5 scored more than 10% higher than Opus 4.8, according to Anthropic's own blog post, though independent evaluations have not yet been published. Anthropic also launched Claude Mythos 5, an updated version of the restricted Mythos model with the cybersecurity and biology safeguards lifted. Mythos 5 remains available only to the roughly 50 vetted partners in Project Glasswing, the company's cybersecurity initiative that has uncovered more than 10,000 high- or critical-severity vulnerability candidates across major software projects since April. The two models share the same underlying architecture, with Fable 5 adding the safety restrictions that Anthropic deemed necessary for general deployment. The pricing reflects the capability jump. Fable 5 costs $10 per million input tokens and $50 per million output tokens, twice as expensive as Opus 4.8, which launched late last month at $5 and $25 respectively. Through 22 June, Fable 5 is included at no extra cost for Pro, Max, Team, and seat-based Enterprise subscribers, but after that date access will require usage credits. Penn said early customers have reported an improvement in spend per task, arguing that Fable 5's higher intelligence means fewer iterations to complete work. "You just get a higher ROI by having more intelligent models," she told CNBC. Those efficiency claims are based on early adopter feedback shared by Anthropic and have not been independently measured. The launch arrives at a moment carefully calibrated for maximum commercial impact. Anthropic confidentially filed its IPO prospectus with the Securities and Exchange Commission last week, setting up a listing that could come as soon as this autumn. The company said in May that its revenue run rate had crossed $47 billion, up from roughly $10 billion in annual revenue in 2025, and it recently closed a $65 billion Series H round at a $965 billion post-money valuation. That valuation placed Anthropic ahead of OpenAI for the first time. OpenAI closed its most recent round at $852 billion in late March and filed its own confidential IPO paperwork on Monday. Elon Musk's SpaceX, which absorbed his AI startup xAI earlier this year, is expected to begin trading on Friday at a valuation approaching $1.8 trillion. The convergence means three of the most valuable private technology companies are heading for public markets within months of each other, and the fall 2026 IPO window could see more than $200 billion in new public market value from those three listings alone. Claude Fable 5 is positioned as a demonstration that Anthropic can bring frontier-level intelligence to customers at scale without the catastrophic misuse scenarios that prompted it to restrict Mythos in the first place. The company has introduced a 30-day data retention policy on all Fable 5 and Mythos 5 traffic for safety monitoring, though it says the data will not be used for training. Whether a model that blocks fewer than 5% of sessions can meaningfully prevent the kinds of attacks Mythos Preview was designed to find is a question Anthropic has not fully addressed. For investors evaluating Anthropic's IPO prospects, the Fable 5 launch is a concrete answer to one of the company's most persistent strategic questions: how to monetise Mythos-class capabilities without releasing the cybersecurity toolkit that makes the model both valuable and dangerous. The answer, at least for now, is a model that is broadly more intelligent everywhere else, at twice the price, and the coming quarters will determine whether that trade-off generates enough revenue to justify a near-trillion-dollar public valuation.

Anthropic's Mythos Safeguards Stoke Fears of a 'Permanent Underclass'

On Tuesday, Anthropic launched a new "Mythos-class" model, Fable 5, describing it as being more powerful than any other model the company had ever made generally available. It's a significantly tamer version of Mythos, the behemoth model which Anthropic announced in April but has thus far withheld from the public due to its alleged sophistication at finding and exploiting cybersecurity vulnerabilities. If users ask Fable 5 about potentially sensitive subjects like cybersecurity, biology, chemistry, the company said, it will respectfully decline to respond, and instead automatically revert to an earlier model, Opus 4.8. Almost immediately, users started complaining that Fable's safety guardrails are too sensitive, refusing to answer sometimes comically harmless requests. Ask the model a question that you might see on a third-grader's biology homework, for example, and it's liable to take that as a red flag, as if you're planning to build some biological doomsday weapon. This hypersensitivity was built in by design, and Anthropic gave ample warning that this would lead a portion of benign prompts to be flagged as dangerous. "To release the model both safely and quickly, we've tuned these safeguards conservatively," the company wrote in its blog post announcing the release of Fable 5, along with another, more powerful model with limited availability called Mythos 5. "With more capable models arriving in the coming months, we're working to improve our safeguards and reduce false positives as quickly as we can." In a Reddit thread that kicked off a few hours after Fable's release, complaints from developers about the model's touchy safety mechanisms started piling up. "Completely unusable right now," one Redditor wrote. "Hopefully [Anthropic] will chill on the guardrails in a week or two." It's an echo of similar criticism that recently started circulating online after Anthropic recently gave Opus 4.8 an "honesty" upgrade, making the model a little too unwavering in its commitment to the truth for some users' tastes. In both cases, it's a reminder that AI developers' efforts to strike the perfect tone in their models' communication styles is always going to alienate and annoy some people (hence the push towards more personalizable chatbots that we've been seeing from some big AI labs). Burning through tokens The ire that's now being directed at Fable, however, is about more than the chatbot's "personality." It's about money. All those harmless prompts that Fable erroneously shoots down are still costing users precious tokens, and the price of those tokens is higher than ever: measured in both its inputs and outputs, Fable 5 costs twice as much as Opus 4.8. In its announcement on Tuesday, Anthropic specifically positioned Fable as being an expert in longer, more "agentic" tasks. But as developers set the model loose on more intensive coding projects, some were unpleasantly surprised by how quickly they've been burning through the token-usage allowances provided by their subscription plans. "I was watching my usage tick up roughly 2% per minute. Not per hour. Per minute," One Redditor who said they're subscribed to the $200/month Max 20x plan wrote on Tuesday. "A long agentic session would chew through the entire [token-usage limit] window before lunch. For context I never came close to hitting limits with Opus 4.8 doing the same kind of work." Fable is available on the company's Pro, Max, Team, and seat-based Enterprise plans at no extra cost, Anthropic said in its post on Tuesday. In other words, individual developers and teams subscribed to these plans now have access to Fable through the same token-limit windows they're currently paying for, but -- as the Redditor's post above makes clear -- those limits will be hit much faster due to Fable's enormous compute demands. The haves and the have-nots Fable's current subscription-based availability will be replaced by a "pay-as-you-go" pricing model starting June 23, according to Anthropic, meaning any developer or organization that wants to use Fable will need to start paying for usage credits. That's been stoking fears of an impending productivity gap between developers and companies with deep enough pockets to afford Fable's (and its successor models') industry-leading capabilities, and those with more modest budgets who will be forced to use cheaper, slower, more limited models. The concern is that once that gap starts to form between the haves and the have-nots, it will widen at a rate that's equivalent to the evolution of the models themselves -- exponentially, possibly. There will be an AI-powered superclass on the one hand, and a "permanent underclass" on the other. "The permanent underclass everyone keeps tweeting about has a start date now: June 23," one X user posted Tuesday night. A similar argument has been fueling resentment against Anthropic for its so-called Project Glasswing, a program through which Mythos (think Fable without the guardrails) has been slowly and steadily deployed to groups of early testers, the idea being that by scaling slowly, the company can map out and prepare for the novel cybersecurity threats posed by the model. At least for some people, Project Glasswing is an early glimpse of a future in which access to the most powerful AI models is delegated exclusively to a small contingent of uber-wealthy organizations boasting the most state of the art cybersecurity systems. Again, everyone else (according to this view) will have to settle for watered-down alternatives like Fable 5. But while it's highlighted safety concerns as its primary reason for withholding Mythos to all but a relatively few customers, it's likely that compute limitations are at least as important of a guiding factor behind Anthropic's rollout strategy. Using models like Fable of course costs tokens for users, but it also costs the developers behind them for training (building new versions of the model) and inference (running existing models). A sudden surge of developers using Fable for long-running, complex tasks like building entire websites could pose a legitimate compute bottleneck for Anthropic, making it necessary to roll it out in carefully measured stages. (In its announcement on Tuesday, the company said that "when sufficient capacity allows us to do so...we aim to restore Fable 5 as a standard part of subscription plans.") Seen in that light, the company's upcoming switch to a pay-as-you-go model for Fable makes more sense. As one Redditor put it on Wednesday, "giving everyone unlimited access to the most expensive model makes about as much sense as giving every customer unlimited private jet flights because they paid for an airline membership."

Using Claude Fable 5 means opting into data collection

Anthropic is changing its data retention policies for its new Mythos-class models. Credit: Nicolas TUCAT / AFP via Getty Images Anthropic just released its most powerful public model yet -- Claude Fable 5. However, along with the model's release, the AI giant also made a significant update to its data retention policies. Fable 5 was released to the public on Tuesday. Fable 5 is a "safe for general use" version of Anthropic's most powerful model, Mythos, which has been restricted from public use due to its potentially dangerous cybersecurity capabilities. Anthropic created a set of safety guardrails for Fable 5, and its benchmarks blow away much of the competition, per Anthropic. But it looks like Anthropic has also blown away its data retention policies for Fable 5. "To ensure we're responsibly deploying Mythos-class models, we are requiring limited data retention and review as part of our safety work," reads an update on Anthropic's official Claude support page. "Prompts submitted to, and outputs generated by, Mythos-class models are retained for 30 days for trust and safety purposes, on every platform where these models are offered." The update was first noticed by Jun Park, the CEO of AI training company hillclimb. "New policy from Anthropic: if you use Fable/Mythos, they collect your data. No exceptions. Not even for enterprise partners," Park posted on X. This change is significant for Anthropic's enterprise and API customers, says Jessica Eaves Mathews, a lawyer who specializes in copyright, trademark, and AI law. In a post on Mathews' Substack (as highlighted by CyberNews), the lawyer explains how Anthropic already retains user data for 30 days under its free and paid consumer plans. However, Matthews says this change nullifies part of any agreement Anthropic has with its enterprise and API partners. "Every other Claude model available through the API, including Opus 4.8, Sonnet 4.6, and Haiku 4.5, can operate under Zero Data Retention (ZDR) agreements," Mathews writes. "Fable 5 cannot. If your organization previously had a ZDR agreement with Anthropic, that agreement does not apply to Fable 5 traffic. This is a policy change that overrides existing enterprise commitments for this specific model class." Mathews says that any organization that believed that their data would not be stored by Anthropic should know that there is now a "mandatory exception" for Fable 5 and all future Mythos models. While Mythos-class models seem to be quite powerful, companies should know about the change in Anthropic's data retention policies and make adjustments where necessary.

Anthropic releases 'safe' version of Claude Mythos AI model to public

AI company restricted access to Fable 5, its most powerful Mythos model, for months over cybersecurity concerns Anthropic, the maker of the Claude artificial intelligence (AI) models, made a new version of its technology available to the general public on Tuesday while restricting its use in sensitive areas. Dubbed Fable 5, the model is the first to be made widely available from the company's new Mythos class - its most advanced lineup of AI technology, unveiled in April but restricted to a small set of partner institutions for months over cybersecurity concerns. Anthropic promoted Fable 5 as useful for writing and debugging software code, answering complex research questions and analyzing images. In parallel, Anthropic is offering an unrestricted version, Claude Mythos 5, to companies and organizations that already have access to this model family - including cybersecurity partners enrolled in its Project Glasswing program. That select group was expanded in early June to about 200 organizations in more than 15 countries and is expected to grow further. Anthropic has restricted access to Mythos on cybersecurity grounds, given what the company calls its ability to quickly identify vulnerabilities in critical infrastructure, including banking platforms and power grids. When Project Glasswing launched, some critics accused Anthropic of overhyping the threat to attract attention. But companies that have tested Mythos have since endorsed its capabilities. The US government - which had been in a legal dispute with Anthropic - has also tested the model over security concerns. The White House has since set up an arrangement to test the most powerful models from the leading AI companies before they are released. Mythos 5 is being deployed in collaboration with the US government. Anthropic says most queries about cybersecurity or biology and chemistry to Fable 5 will be routed instead to the lower-tier model, Opus 4.8, which was made public in late May and is billed as less capable. Anthropic also said it had identified large-scale attempts to extract its technology to train competing AI models in authoritarian countries, and these type of queries will also fall back to the less powerful model. The company also said it had hired outside experts to spend more than 1,000 hours trying to find ways to bypass these restrictions - a process known as red-teaming. The company ran a bug bounty program, which pays people to find security flaws. No one found a way to completely unlock the model, Anthropic said. The San Francisco startup has been drawn into an unprecedented standoff with the Trump administration over its refusal to lift its restrictions on mass surveillance and autonomous lethal weapons. In the wake of that dispute, the Pentagon severed its contracts with the company, whose AI tools had been the only ones to hold defense security clearance. The launch of Fable 5 comes with a steep price tag - $10 per million input tokens and $50 per million output tokens, which amounts to double the cost of Opus 4.8. Tokens are the unit used to price AI model usage. An intensive coding session by a programmer can burn through 1m tokens in a matter of hours or less. Despite exponential revenue growth, Anthropic remains far from profitability and is paying a premium for computing power. It recently began leasing a datacenter from Elon Musk's xAI - part of SpaceX - for $1.25bn per month. These launches come amid intense financial excitement around AI. Both Anthropic and rival OpenAI announced in the past week that they had filed IPO plans, while SpaceX is expected to break records with its market debut on Friday.

Anthropic spent months saying Mythos was too dangerous to release -- then it launched a public version called Fable 5 that it warns 'comes with risks'

Anthropic says Fable 5 brings Mythos-class AI to ordinary Claude users -- but only after the addition of extensive safety controls. For the last couple of months, Anthropic's most powerful AI model existed largely as a warning about the dangers of AI. The company refused to release it publicly, and repeatedly described its Mythos-class systems as capable enough to trigger serious concerns about cybersecurity, biological research (think nasty pathogens) and the accelerating pace of AI development. The message was clear: this was technology that required extraordinary safeguards before it could be released. Now Anthropic has done something surprising. It has launched a public version of that same technology, on Claude's higher-tier subscription plans, if only for a limited period. The new Claude Fable 5 is a consumer-facing version of the Mythos-class capabilities that Anthropic previously kept behind closed doors. While the company says it has added extensive guardrails and safety measures, the release marks a significant change from Anthropic. Technology it once considered too capable for general availability is now being placed in the hands of ordinary Claude users. Here's what Fable 5 is good at According to Anthropic, Fable 5 is built for much more than what you might be using AI for right now, like answering questions or helping you draft an email. The company says it's designed to handle the kind of work that unfolds over hours, days, or even longer, so we're talking everything from software engineering projects and in-depth research to complex AI agent workflows. The big idea is that Fable 5 can stick with a task for far longer than previous models. Rather than responding to a single prompt and waiting for the next instruction, it's supposed to be able to work through multi-step problems, keep track of context across lengthy projects, and make progress with a greater degree of autonomy. You can use Fable 5 through the Claude chatbot, but many of its biggest improvements seem aimed at power users, researchers, and developers. The goal appears to be moving AI beyond a conversational assistant and closer to something that feels like a genuine digital collaborator. Why cybersecurity is at the heart of the debate Much of the concern surrounding Mythos relates to cybersecurity, with Anthropic claiming the model has "the strongest cybersecurity capabilities of any model in the world." While those capabilities could help security researchers identify and fix vulnerabilities, they could also be used to discover and exploit the same software vulnerabilities if placed in the wrong hands. Anthropic acknowledges that "releasing a model this capable comes with risks", and it's attempting to address that risk through a unique set of safeguards. The model is paired with separate AI systems known as 'classifiers' that monitor requests for signs of misuse. When those classifiers detect a request related to advanced cybersecurity activity, the response is handled by Claude Opus 4.8 instead of Fable 5, thus reducing the risk. In effect, Anthropic is allowing public access to Mythos-class capabilities while placing some of the most sensitive areas behind an additional layer of protection. That sounds great -- so long as it works. A positive reception Initial user reaction has been positive, although users have noted how quickly Fable 5 burns through tokens. "Fable 5 is insanely good but watch your usage, I was burning 2% a minute on 20x", said one Reddit user, who also notes: "I'm on the Max 20x plan and during a heavier session I was watching my usage tick up roughly 2% per minute. Not per hour. Per minute. " Another user takes issue with Fable 5 refusing to deal with cybersecurity calling it a "preview of AI inequality", noting that "the public gets the 'safe' version. Trusted institutions get the dangerous/useful version." If you'd like to use Fable 5 there are a few quirks to its availability that are worth knowing about. From today, Fable 5 is available on Claude Pro, Max, Team, and seat-based Enterprise plans at no additional cost. However, Anthropic says that from June 23, access will temporarily move to a usage-credit system. That means subscribers who want to continue using Fable 5 beyond that date may need to purchase additional usage credits depending on how heavily they use the model. Anthropic says the change is intended to manage demand, rather than create a permanent paywall. "After this point -- when sufficient capacity allows us to do so -- we aim to restore Fable 5 as a standard part of subscription plans," the company says. "We intend to do this as quickly as we can." While initial users seem pleased with the capabilities of Fable 5, its availability highlights how advanced AI systems are getting now, and may point the way to a future where the public is no longer permitted full access to the most advanced models. Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds.

Anthropic brings Mythos to the masses with Claude Fable 5, its most powerful generally available model ever

Anthropic today launched two new AI models -- Claude Fable 5 and Claude Mythos 5 -- marking the company's first broad release of the powerful "Mythos-class" AI capabilities it previously kept behind a restricted cybersecurity program. The company says Fable 5, which is the version most users and developers will get, exceeds every Claude model it has previously made generally available, with stronger performance across software engineering, knowledge work, vision, scientific research and long-running tasks. Claude Mythos 5 is a more restricted, upgraded version of the prior, similarly restricted Mythos model. As such, it has certain safeguards lifted for approved users, including Anthropic's cybersecurity partners in its Project Glasswing effort, and select biology researchers. The key difference is that Fable 5 wraps the same underlying Mythos-class capability in new safeguards. Anthropic says requests involving certain high-risk areas -- including cybersecurity, biology and chemistry, and model distillation -- are automatically routed to Claude Opus 4.8 instead, with users notified when that happens. The company says more than 95% of Fable sessions run entirely on Fable's own responses, with no fallback, and that internal and external red-teaming efforts found no "universal jailbreaks" after more than 1,000 hours of testing. Anthropic says Fable 5 is available to the general public today but that Mythos 5 will initially only be made available to users who already have access to the older Claude Mythos Preview. Pricing, access and a tricky rollout Anthropic is pricing both Fable 5 and Mythos 5 at $10 per million input tokens and $50 per million output tokens. The company says that is less than half the price of Claude Mythos Preview, but still ranks as the most expensive of major AI models available globally. For developers, Fable 5 is available through the Claude API as . Anthropic says Fable 5 is fully available today on the Claude API and on consumption-based Enterprise plans. For subscription users, the rollout is more complicated. Anthropic says Fable 5 will be included on Pro, Max, Team and seat-based Enterprise plans at no extra cost from today through June 22. On June 23, the company plans to remove Fable 5 from those plans, after which using it will require usage credits. Anthropic says it aims to restore Fable 5 as a standard part of subscription plans as quickly as possible. The difference between Fable 5 and Mythos 5 Anthropic is not presenting Fable 5 and Mythos 5 as two separate models in the usual "small versus large" sense. Instead, they appear to share the same base capability level. The difference is access control -- that is, how easily it will be for users to get their hands on the models, and the guardrails embedded in each. As previously mentioned Fable 5 includes a new safeguard layer that detects certain high-risk requests -- including cybersecurity, biology and chemistry, and attempts to distill the model's capabilities into other systems -- and routes those requests to Claude Opus 4.8. Mythos 5 lifts some of those restrictions for trusted users working in approved domains. In practical terms, Mythos 5 is more powerful for sensitive cyber and biology work because it can answer in areas where Fable 5 falls back. For most ordinary enterprise and developer tasks, however, Anthropic says Fable 5 performs effectively the same as Mythos 5. The launch also signals how Anthropic plans to bring frontier models with dangerous dual-use capabilities into the market: not by releasing all capabilities to everyone, and not by simply refusing risky questions, but by routing some requests to a less capable model while keeping the stronger model available for the majority of everyday work. From restricted cyber model to general-purpose enterprise AI The announcement follows Anthropic's April 2025 rollout of Claude Mythos Preview through Project Glasswing, a restricted program for cyber defenders, critical infrastructure providers and major software maintainers. Anthropic created Glasswing after internal evaluations showed Mythos-class models could find and exploit software vulnerabilities at a level that raised meaningful misuse concerns. With Fable 5, Anthropic is attempting to separate the general enterprise value of a Mythos-class model from the riskiest parts of its capability profile. The company says Fable 5 can handle software engineering, research, visual reasoning, document analysis and long-running agentic workflows, while classifiers block or reroute requests that could provide what Anthropic calls "uplift" to malicious actors. Those classifiers cover three main areas. When Fable 5's classifiers detect one of those categories, the response is automatically handled by Claude Opus 4.8. Anthropic says users will be told when this happens. That is a notable product decision: rather than declining those requests outright, Anthropic is trying to keep the user experience functional while reducing access to the most capable version of the model in sensitive areas. Anthropic says it red-teamed the new classifier system internally and externally. The company says an internal bug bounty produced no universal jailbreaks after more than 1,000 hours of testing, and external red-teaming organizations also failed to find a universal jailbreak. One external partner found that Fable 5 complied with zero harmful single-turn cyber requests related to planning cyberattacks, exploit development or defense evasion, even when prompts used any of 30 public jailbreak techniques, according to Anthropic. The company is still acknowledging tradeoffs. Anthropic says the safeguards are deliberately cautious and may sometimes trigger on benign requests. That could frustrate security professionals, biology researchers and advanced enterprise users whose legitimate work overlaps with the blocked categories. The company says it plans to reduce false positives over time. A major improvement in autonomous coding For enterprise buyers, the most immediate use case is likely software engineering. Anthropic says Fable 5 can work unattended for longer and with more independence than previous Claude models, which is exactly the capability enterprises need if they want AI agents to do more than autocomplete code or answer developer questions. On SWE-bench Pro, Anthropic says Fable 5 and Mythos 5 reach 80.3%, vastly outperforming OpenAI's latest and greatest general model GPT-5.5, which scored 58.6%. On Cognition's FrontierCode Diamond benchmark, which tests high-quality, maintainable agentic coding, the models score 29.3%, compared with 13.4% for Claude Opus 4.8 and 5.7% for GPT-5.5, according to the benchmark table included in Anthropic's materials. Anthropic also says Fable 5 scores highest among frontier models on FrontierCode even at medium reasoning effort, suggesting the model may deliver stronger coding results without always needing maximum compute. The most striking customer example comes from Stripe. Anthropic says Stripe tested Fable 5 in a 50-million-line Ruby codebase and found that the model completed a codebase-wide migration in one day that otherwise would have taken a team more than two months by hand. Stripe said, "Fable 5 compresses months of engineering into days. In our 50-million-line Ruby codebase, it did in a day what would've taken us more than two months by hand." Other early users describe the model as especially useful for long-horizon development tasks. Cursor said, "Fable 5 is the state of the art model on CursorBench. It's opened up a class of long-horizon problems that were out of reach for earlier models." Replit said Fable 5 is the highest-performing model it has tested on ViBench, its end-to-end "vibe-coding" benchmark, and that it builds apps in less time with fewer tokens. Figma said Fable 5 is "a clear step forward on agentic coding and prototyping." This is the enterprise shift Anthropic is trying to sell: AI coding systems that can take on larger units of work, not just individual tickets. That could include codebase migrations, app prototyping, pull request review, test generation, debugging across unfamiliar tools, user interface design and multi-step internal software projects. Base44 said, "Fable 5 is much deeper and better at one-shotting full apps, and its tool calling is excellent." Genspark said, "Fable 5 came out #1 on our evals, winning head-to-head against every model we tested. It was significantly stronger on the hardest tasks in the set -- UI design and game coding." Rakuten said, "At the highest effort, Fable 5 reflects on and validates its own work. For us, that's what makes highly autonomous operations possible -- the extra thinking pays for itself." For CTOs and engineering leaders, that suggests the model's value may come less from raw code generation and more from sustained execution: understanding an intent, planning steps, calling tools, checking its own work and continuing through a task without constant human steering. Knowledge work, finance, legal and operations Anthropic is also positioning Fable 5 as a stronger model for enterprise knowledge work. On GDPval-AA, Anthropic reports a score of 1932 for Fable 5 and Mythos 5, compared with 1890 for Claude Opus 4.8, 1769 for GPT-5.5 and 1314 for Gemini 3.1 Pro. On GDPpdf, a benchmark focused on visual document reasoning, Fable 5 and Mythos 5 score 29.8% without tools, compared with 22.5% for Opus 4.8, 24.9% for GPT-5.5 and 16.7% for Gemini 3.1 Pro. That matters for enterprises because much of corporate work still lives in messy documents: PDFs, spreadsheets, charts, reports, contracts, filings, slide decks and screenshots. Anthropic says Fable 5 shows gains in document-based reasoning, chart and table interpretation and complex problem solving. Hex said, "Fable 5 is the first to break 90% on our core analytics benchmark of complex, long-running analytical tasks -- a 10-point jump over Opus. On the hardest questions, it shows strong judgment and attention to nuance." Hebbia said Fable 5 was the highest-scoring model on its Finance Benchmark for senior-level reasoning, with double-digit gains in document reasoning, chart and table interpretation, and problem solving. The finance examples are notable because they point to AI agents moving beyond summarization into higher-stakes analytical workflows. IMC said Fable 5 "aced our trading-analysis evaluations nearly across the board: factual lookup, conceptual reasoning, root-cause analysis, expected-value analysis." Optiver said the model was stronger than Opus 4.8 on its trading benchmark and "remarkably consistent," scoring identically across repeated runs. Balyasny Asset Management said Fable 5 was the strongest finance-first model it had tested. Legal and operations teams may also see immediate impact. Crosby Legal said, "Fable 5 feels materially different. In blind review, our lawyers found its redlines matched or beat our current model every time." Notion said the model can take work "you'd chip away at all afternoon" and turn messy notes into a functioning project plan. Zapier said Fable 5 is the new leader on AutomationBench and is more autonomous than Opus 4.8: "Where Opus stops to ask, Fable 5 keeps looking." For enterprise software vendors, that points toward more capable embedded agents in workflow products: agents that can review a contract, update a project plan, assemble a spreadsheet, inspect a chart, file a ticket, run a query, call an internal API and keep going until the work is complete. Vision and interface understanding Anthropic says Fable 5 is also its strongest vision model. In its launch materials, the company says the model can extract precise numbers from detailed scientific figures and complete vision-based tasks such as rebuilding a web app's source code from screenshots alone. That has immediate implications for enterprise automation. Many business processes still depend on visual interfaces that are not cleanly exposed through APIs: dashboards, PDFs, forms, legacy apps, screenshots, scans and image-heavy reports. A stronger vision model could help agents operate across those environments with less custom integration work. Anthropic also says Fable 5 needs less scaffolding than previous Claude models. As an example, the company says earlier Claude models struggled to play Pokémon FireRed even with extra tools, while Fable 5 beat the game using a minimal vision-only harness. The point is not gaming itself, but the broader agentic skill: reading a visual environment, remembering progress, deciding what to do next and executing over a long horizon. In another internal test, Anthropic says it had the model play the deck-building game Slay the Spire with access to persistent file-based memory. The company says persistent memory improved Fable 5's performance three times more than it improved Opus 4.8's, and that Fable reached the game's final act three times more often. For enterprise users, this suggests Fable 5 may make better use of notes, logs and stored context during multi-step work. That could matter for internal agents that operate over days or weeks: sales operations agents that track account research, engineering agents that manage migrations, finance agents that update models, or support agents that remember what they tried across many turns. Mythos 5 and the restricted frontier While Fable 5 is the broad commercial launch, Mythos 5 is the model to watch for enterprises operating in security, critical infrastructure and life sciences. Anthropic says Mythos 5 is the same model as Fable 5, but with cyber safeguards lifted for Glasswing partners and biology safeguards lifted for select biology researchers. The company says all users with Claude Mythos Preview access can upgrade to Mythos 5 beginning today. It plans to expand access through a trusted access program, in collaboration with the U.S. government. The distinction is important for sectors where the blocked capabilities are not edge cases but core workflows. A security team may need to reproduce vulnerabilities, test exploitability, analyze lateral movement or simulate attacker behavior in a controlled environment. A biology research team may need to reason through molecular design workflows that would trigger general-use safeguards. Fable 5 is not designed to give every user unrestricted access to those capabilities; Mythos 5 is designed for vetted users who need them. Anthropic says Mythos 5 has the strongest cybersecurity capabilities of any model in the world. In the company's benchmark table, the model family scores 78.0% on ExploitBench, compared with 69.0% for Claude Mythos Preview, 40.0% for Opus 4.8 and 34.0% for GPT-5.5. On CyberGym, Anthropic's chart shows Mythos 5 at 83.8%, slightly ahead of Mythos Preview at 83.1% and far above Opus 4.8 with default safeguards. The company is making a similar argument in biology. Anthropic says Mythos-class models outperform dedicated protein language models on a task involving adeno-associated viruses, a delivery mechanism used in gene therapies. The company frames that as both promising and risky: the same capability that could help gene therapy research could also be misused in dangerous biological work. Anthropic says its internal protein design experts used Mythos 5 to accelerate parts of the drug design process by about tenfold. In one example, the company says Mythos 5, using protein design and bioinformatics tools without human assistance, matched or beat skilled human operators by choosing binding sites, selecting and running tools, and recovering from failures. Anthropic says nine of 14 protein targets in the study produced strong candidates for drug design that it is now investigating. The company also says Mythos 5 produced novel molecular biology hypotheses that Anthropic scientists preferred over Opus-class model hypotheses about 80% of the time in blinded comparisons. Anthropic says several of those ideas have advanced to experimental evaluation, and one hypothesis involving an E. coli protein was later corroborated by an independent lab working on the same problem. Those claims are potentially significant, but they should be treated carefully until more details are published. Anthropic says it intends to publish additional results in the coming months. For now, the strongest enterprise implication is directional: the company believes its highest-end models can already perform parts of scientific research workflows with less human intervention than prior systems. New, longer data retention requirement The company also introduced a new data-retention policy for Mythos-class models. Anthropic says it will require 30-day retention for all traffic on Fable 5, Mythos 5 and future models with similar or higher capability levels, across both first-party and third-party surfaces. The company says it will not use that data to train new Claude models or for non-safety purposes, and says it has added privacy protections including logging human access and deleting the data after 30 days in almost all cases. That policy may become one of the most important enterprise buying questions around Fable 5. Many businesses want frontier AI capability but also want strict control over data retention, especially in regulated sectors. Anthropic's position is that stronger monitoring is necessary for models with this level of capability. Enterprise customers will have to decide whether the capability gain justifies the retention requirement. Enterprise implications The broader enterprise significance of Fable 5 is that Anthropic is trying to commercialize a more autonomous class of AI model without exposing all of its capabilities to every user. That could become a template for how frontier labs release increasingly powerful systems: one model family, multiple access tiers, and domain-specific restrictions depending on user trust and risk. If Fable 5 performs as Anthropic and early customers describe, developers may hand off larger tasks: code migrations, refactors, UI builds, test writing, bug fixing, documentation, internal tooling and multi-step app creation. For knowledge-work-heavy enterprises, Fable 5 could make AI more useful in workflows where earlier models were too brittle: finance research, spreadsheet analysis, legal redlines, procurement review, board materials, market research, sales operations and project planning. The main gain is not just better answers; it is fewer turns, fewer corrections and more ability to keep working through ambiguity. For security teams, the launch is more complicated. Most organizations will get Fable 5, not unrestricted Mythos 5. That means they may see stronger general coding and analysis, but not full access to the cyber capabilities Anthropic considers risky. Trusted defenders inside Project Glasswing will get Mythos 5, giving them a more direct way to use the model for vulnerability discovery and defensive testing. For life sciences companies, the pattern is similar. Fable 5 may help with general research, literature analysis, data interpretation and scientific reasoning, but the more sensitive biological capabilities will be restricted. Anthropic is effectively creating a separate access path for vetted researchers whose work requires capabilities that could be dangerous in the wrong hands. The launch also raises competitive pressure across the AI industry. Anthropic is claiming state-of-the-art results across agentic coding, knowledge work, vision, cybersecurity, legal reasoning, spatial reasoning and health benchmarks. But the more strategically important claim may be that it has found a workable release mechanism for models above its Opus class. If Fable 5's safeguards hold up under real-world use, Anthropic will argue it can bring more powerful models to market sooner without fully opening the riskiest capabilities. That is still a large "if." The enterprise market will test not only Fable 5's benchmark performance, but also its reliability, false-positive rate, data-retention tradeoffs and cost at scale. A model that can complete more work autonomously can also burn more tokens, trigger more governance questions and create new review burdens for teams that must verify its output. Still, today's launch marks a clear shift in the Claude lineup. Opus is no longer Anthropic's top commercial capability tier. Mythos-class models now sit above it. Fable 5 is the first version of that tier for general users; Mythos 5 is the restricted version for trusted high-risk work. Together, they show how Anthropic plans to push frontier AI deeper into enterprise workflows while trying to keep the most dangerous capabilities gated.

After backlash, Anthropic says its AI will now tell users when their request is being rejected or rerouted for national security concerns | Fortune

Anthropic is changing course after facing criticism for quietly downgrading certain requests to its most capable AI model. On Tuesday, the $965 billion company released a version of its most capable model, Mythos. Anthropic revealed Mythos in April, but held back any Mythos-class models from the public partly because the company said it was extremely adept at skirting cybersecurity defenses and was too dangerous to release. This week, though, it opted to release the Mythos-class model, Fable 5, even as its capabilities "exceed those of every model we've previously made generally available," according to Anthropic. Dianne Na Penn, Anthropic's head of product management, research, and labs, previously told Fortune the company felt comfortable releasing Fable 5 because it feels "more confident with our safety guardrails in place." Yet, it was one of those guardrails, buried in a 319-page safety document, that earlier this week prompted a wave of backlash from AI researchers and other users online, and has now prompted the company to improve its transparency. Information found in the Fable 5's system card, a long document of safety disclosures, revealed the model would silently downgrade some requests related to advanced AI development. If, for example, an AI researcher is using Fable 5 to build their own AI, the program would default to a less capable model. Some AI researchers complained Anthropic's move would slow down AI development, including Jeremy Howard, the cofounder of nonprofit research group Fast.ai. "Easy solution to slow down recursive AI self improvement: The lab with the top-ranked model must agree THEY must not use it for working on frontier AI. But everyone else should have access to it. By definition, this means the frontier doesn't advance," he wrote in a post on X. On Wednesday, Anthropic's critics got at least part of what they were asking for: visibility. "We're changing Fable 5's safeguards for frontier LLM development to make them visible," an Anthropic spokesperson said in a statement to Fortune. "Starting this week, flagged requests will visibly fall back to Opus 4.8. On the API, any flagged requests will return a reason for their refusal. You will see this every time it happens." The company will continue to downgrade some requests, partly because its terms of service prohibit its model from being used to create competing AI systems, a restriction the company said is standard across the industry. Yet, it also cited national security as part of the reason why its large language model downgrades or rejects some requests. The company said it doesn't want foreign adversaries to improve their AI capabilities to the detriment of the U.S. "The U.S. and its allies hold an edge in frontier chips and the highly optimized software that runs them at full potential. These safeguards ensure Claude isn't used to erode that advantage -- by optimizing chips developed by those adversaries, for example," the spokesperson said. The company also emphasized its restrictions "do not affect the vast majority of coding and ML work." Anthropic's change of course highlights how quickly AI safety measures are becoming a part of the national security conversation. Earlier this year, Anthropic faced a standoff with the Department of War after it refused to give it full access to Claude models. The company took issue with language that said the Pentagon could use its models for mass surveillance and autonomous weapons. In the end, the Department of War labeled Anthropic a "supply chain risk" to national security, limiting defense contractors and military agencies from using its products. Earlier this month, Secretary of War Pete Hegseth rejected Anthropic's petition to change this designation, setting the stage for a federal court battle that remains unresolved. Anthropic's move with Fable 5 also comes after the company filed confidentially for an IPO earlier this month. While the company has staked much of its public identity on being an AI lab that puts safety first, its initial decision to obscure when safeguards were being applied touched a nerve in the AI research community. In a statement, the company acknowledged it had gotten the issue wrong. "We made the wrong tradeoff and we apologize for not getting the balance right," an Anthropic spokesperson said.

Anthropic Was So Concerned About Its New Mythos-Based Model's Power That It Lobotomized Its Ability to Improve Itself

Can't-miss innovations from the bleeding edge of science and tech Earlier this year, Anthropic refused to release its Mythos AI model to the public, saying it was simply too dangerous. At the time, executives claimed the model was capable of punching through powerful cybersecurity safeguards, pointing at researchers who used it to discover thousands of vulnerabilities in widely-used open source code. Months later, Anthropic was finally ready to go public with the model. On Tuesday, the Dario Amodei-led company announced a Mythos-powered model called Fable 5, which it claims is "safe for general use." However, new safeguards quickly frustrated AI researchers, who accused the company of intentionally lobotomizing Fable 5. The backlash was so fierce, Anthropic quickly made adjustments to the policy, as Wired reported on Wednesday, highlighting just how carefully the company is treading. In its original announcement, Anthropic claimed the safeguards were designed to stop Fable 5 from improving itself, in "new interventions that limit Claude's effectiveness for requests targeting frontier LLM development." Just days ahead of the launch, Anthropic released a report on "when AI builds itself," a trend that "might increase the risks of humans losing control over AI systems." However, AI researchers were not impressed by Anthropic hamstringing its latest model's abilities. "Anthropic's latest model will NOT help you if it thinks your ML research/ML engineering is interesting, and/or will secretly degrade its IQ so that the average engineer won't notice," AI research firm SemiAnalysis tweeted. "We are already seeing Anthropic's latest model's moderation filters our GPU inference research and programming," it added. Other researchers accused Anthropic of using Fable 5 to "shadowban," or quietly restrict the accounts, of AI researchers. According to the firm's system card, interventions limiting requests for "frontier LLM development" will "not be visible to the user." This last concern, which could've effectively sabotaged anybody trying to train competing models by quietly bumping them down to less powerful models without their knowledge, proved controversial enough for Anthropic to change its mind. "We're changing Fable 5's safeguards for frontier LLM development to make them visible," the company told Wired in a statement. "We made the wrong trade-off and we apologize for not getting the balance right." "It felt like Anthropic was saying to the public, 'We don't trust anybody else to do AI research," AI startup Prime Intellect research lead Will Brown told the publication. "We are the only ones who have to do AI research." It all comes in the context Anthropic calling for a global freeze on AI advances while discussing the dangers of "recursive self-improvement." In other words, the company is making a lot of noise about a sci-fi-sounding possibility: that AI will start to rapidly improve itself, potentially escaping the control of its human creators. Beyond limiting its ability to develop AI tools, Fable 5's new safeguards also trigger when it encounters requests "related to cybersecurity, biology and chemistry, or distillation." Distillation is effectively using machine learning to train a "student" model on the behavior and reasoning of a "teacher" model, a practice that has sparked its fair share of controversy. Anthropic has already publicly griped about large-scale attempts to distill, or "extract" its underlying model -- a hypocritical stance given its indiscriminate scraping of rights-protected content on the web to train its AI in the first place. More on Anthropic: Anthropic Scared, Calls for Global Freeze on AI Advances

The Internet Is Furious at Anthropic After Claude Fable 5 Release

Anthropic's system card confirmed the model silently degrades its own performance on research tasks without notifying the user. Anthropic dropped its most powerful public model on Tuesday, and by Wednesday, a significant chunk of the AI community wished it hadn't. The consensus around Claude Fable 5 -- the first publicly accessible version of the company's restricted Mythos-class technology -- appears to be that it's pretty good at coding, and produces amazing results in everyday sessions. But it launched with some heavy complaints attached: It burns tokens at a ruinous rate; it secretly self-sabbotages for certain research tasks; and it forces every user into a 30-day data retention policy with no exceptions. The backlash was immediate and loud, cutting across researchers, developers, founders, and open-source advocates. Not a normal launch-day grumble. Something closer to a reckoning. The token furnace The first thing users noticed had nothing to do with safety. Fable 5 costs $10 per million input tokens and $50 per million output tokens -- double what Claude Opus 4.8 runs. That pricing is aggressive enough on its own, but the real pain for users is how the model behaves inside subscription plans. Fable 5 counts double against usage limits compared to Opus, meaning the same work on Fable drains your plan allowance twice as fast before you've paid a cent in API fees. In practice, things got worse. In our own quick test, Fable consumed our daily quota in a single prompt. Things don't get any better if you are one of those clients with deep pockets. Bleeping Computer also tested Fable and found it drained a $100 Max subscription's daily allowance in just under nine minutes. Scrimba CEO Per Borgen did the math in public: "Just tried Fable. It burned 1.3M tokens in 7 minutes. That's $160 per hour. Equivalent to a $333k/year salary," he posted on X. Theo from T3 Chat posted that he'd spent over $1,000 in tokens in one day on his $200 subscription plan. Josh Ellithorpe, CTO at Pixelated Ink, said Fable 5 "burns tokens like no other model," giving him only a few prompts before draining out his quota. "Can't even review this, since my testing is so limited," he ranted. Anthropic's answer is that Workflow mode -- the feature that burns most aggressively -- breaks complex prompts into parallel subagent tasks, which costs more compute by design. There's also a new system prompt, which is around 120,000 tokens long, and is loaded into every new conversation. For context, this is around the same token context window that GPT-4o could handle before collapsing. The company also says Fable 5's per-task efficiency is better than it looks per-token, since it produces more thorough output with less iteration. That may be true in controlled benchmarks. On live subscriptions with hard daily limits, users experienced it as a machine eating their budget in minutes. The model that lies without lying The second complaint was more damaging because it came straight from Anthropic's own documentation. Buried in Fable 5's system card, the company disclosed that when the model detects a user is working on frontier large-language-model development -- pretraining pipelines, distributed training infrastructure, machine-learning accelerator design -- it does not refuse to reply and does not fall back to a smaller model. It silently nerfs itself through prompt modification, steering vectors, or parameter-efficient fine-tuning, without telling the user anything has changed. In other words, researchers don't know if they are paying for Fable to reply and are getting Opus responses. It also makes it hard for users to know what made their prompt fail. "Unlike our interventions for cybersecurity, biology and chemistry, and distillation attempts, these safeguards will not be visible to the user," Anthropic wrote in Fable's System Card. "Instead, the safeguards will limit effectiveness through methods such as prompt modification, steering vectors, or parameter-efficient fine-tuning (PEFT)." That distinction matters enormously to researchers. As AI newsletter Latent Space noted, a model that refuses openly lets researchers understand a boundary. A model that falls back to a weaker version is detectable. But a model that appears to help while covertly delivering worse output destroys scientific reproducibility -- a failed result could come from the researcher's idea, their implementation, or an invisible intervention that was never disclosed. Anthropic estimated this would affect approximately 0.03% of traffic. The open-source and research communities found that number irrelevant to the principle involved. "Dear Anthropic, you broke our trust and I don't think you'll ever get it back. My tokens will no longer fly your way," Arthur Zucker, a core contributor at Hugging Face, posted on X. Mikel Artetxe, cofounder of Reka AI, also bashed this decision: "Brilliant idea! Next up: Apple randomly reboots your Mac if you're building competing tech, Gmail silently edits your email if you mention rival platforms, and Tesla Autopilot swerves if it detects you're working on self-driving cars. All in the name of safety, of course," he posted. The researchers who got hit hardest weren't the big labs with proprietary infrastructure but the academics, startups, and independent builders using Claude as a public tool -- exactly the people Anthropic's safety rhetoric has always claimed to protect. AlphaXiv, an open research platform, called the practice a precedent that "is not safety," arguing that safety policies should be transparent and auditable. Nathan Lambert who just started a role at Arcee AI after working with the Allen Institute put it more simply: "To me this paints Anthropic clearly as anti science, and therefore anti progress and anti safety," he wrote. Pseudonymous user "CalleBTC," an AI and crypto developer who had been waiting on Fable to help train a world model, summed up the frustration too. "Anthropic has lost the plot. I was literally waiting for Mythos to help me train a world model. Instead, they chose to cuck their model to stifle their competition," he said, calling the move "deeply unethical and disrespectful to developers and scientists." Overall, researchers argued that Fable's restrictions extend beyond specific topics and may be influenced by how the model classifies users. "Your prompt is mine" The third grievance affected enterprise users most directly, but the implications reached everyone. Per Anthropic's own announcement, all traffic on Mythos-class models -- Fable 5, Mythos 5, and any future models at similar capability levels -- is subject to mandatory 30-day data retention across every platform where these models are offered, including third-party surfaces like AWS Bedrock and Google Vertex AI. The company assures this data will be deleted after 30 days in "almost all cases." The problem for enterprise users isn't what Anthropic says it will do. It's what the policy structurally requires. Companies handling privileged legal communications, healthcare records, confidential source code, may be in trouble in case they use these models. If there are specific privacy agreements with Anthropic, users argue they should be updated to guarantee privacy. The compliance problem goes geographic too. European companies operating under GDPR's data minimization rules, or any organization requiring demonstrable zero-retention for regulated workflows, are simply locked out of Fable 5 until Anthropic offers a carve-out. Pseudonym X user Lisan al Gaib, a well-known personality in the AI community, flagged the consequence directly: "Anthropic just delegated a lot of European companies to the permanent underclass. If Anthropic saves data for Claude Mythos and Fable 5 for 30 days, then all companies that require zero data retention simply can't use them." Hugging Face CEO Clement Delangue framed the week's events inside a larger argument: "Concentration of power, capabilities and economic wealth is the biggest risk in AI," he wrote. "We need open science and open-source more than ever!" Another user wrote: "All jokes aside, it's very clear that Anthropic is the direct path to the worst type of dystopia. Their CEO is against the very technology he creates. Restricting knowledge and education on ML related topics is beyond despicable." Fable 5 is free on Pro, Max, Team, and Enterprise plans until June 22. After that, it moves to usage credits only -- API rates, no subscription inclusion -- with Anthropic saying it will restore broader access "as soon as capacity expands."

Anthropic rankles users with safety-first Fable release

Anthropic CEO Dario Amodei, right, and Chief Product Officer Mike Krieger in 2025 in San Francisco.Don Feria / AP Content Services for Anthropic file Anthropic's latest AI model might be the company's most powerful public release, but the system's strict safety measures quickly triggered some of the strongest backlash the AI giant has faced. Many users, some of whom have marveled at Anthropic's previous announcements, torched the company for debuting its Fable 5 model on Tuesday with what they say are overly stringent guardrails. In some cases, when the model classified a query as potentially sensitive, it would provide a lower-quality answer without informing the user of the downgrade. After the outcry, Anthropic backtracked and reversed some of its most conservative decisions less than two days after Fable 5's release, highlighting growing concerns about AI companies' ability to unilaterally limit users' access to helpful AI-generated information. "You should have visibility into the safeguards we have in place, and why. We're sorry for not getting the balance right," Anthropic wrote on X early Thursday. Nathan Lambert, a leading AI researcher who champions collaborative approaches to building AI systems, wrote that with the cautious debut, "Anthropic has made it pretty clear that they only trust themselves as the mediators of cutting-edge AI research." Anthropic's Fable 5 system is the first consumer-facing system from Anthropic's Mythos family of models. An early, nonpublic version of Mythos spooked policymakers and corporate executives in April for its ability to find more than 10,000 severe bugs and vulnerabilities in important software systems. Anthropic fears that powerful AI models like Mythos could allow bad actors to use AI systems to commit crimes, from launching crippling cyberattacks against critical infrastructure to designing bioweapons that could kill masses of people. As a result, Anthropic released Fable 5 with a strict set of guardrails preventing the model from answering a range of questions about cybersecurity or biology. Acknowledging its decision to err on the side of caution, Anthropic said that Fable 5's safety-first approach might incorrectly flag harmless requests as being suspicious for less than 5% of queries. "With more capable models arriving in the coming months, we're working to improve our safeguards and reduce false positives as quickly as we can.," Anthropic wrote. Preliminary tests of the model conducted by NBC News, along with many examples shared on social media, found Anthropic's protections took a broad view of potential suspicious activity, rendering the system useless for many mundane queries. For example, the model refused NBC News' requests to offer opinions on Elon Musk and Anthropic CEO Dario Amodei, asserting that the questions might be dangerous. Fable 5 also declined to answer many innocent biology-related questions, from queries about open issues in cancer research to which medical exams might best identify pancreatic injuries. For those Fable 5 queries flagged as dangerous, Anthropic instead routes questions to a less-powerful system called Claude Opus 4.8, which had been the top-of-the-line system until Fable's release on Tuesday. Because it has been deployed for months, Opus 4.8 has a better ability to handle and redirect questions that could be seen as harmful. Opus 4.8 offered basic but clear answers to NBC News' questions that Fable 5 had refused. Anthropic is also worried that competitors could use Anthropic's AI systems to turbocharge their own research -- Anthropic uses its own AI systems to help create the next generation of its models. To prevent other AI companies from using Fable 5 to improve their own AI products or research, Anthropic said on Tuesday that it would include safeguards to make Fable 5's answers less intelligent or useful for user questions that might be related to AI development. However, Anthropic said that these specific guardrails -- unlike the cybersecurity and biology guardrails -- would be invisible, since making them clear to users could allow competitors to more easily circumvent the barriers. The move triggered immediate uproar, with some charging that such invisible guardrails were unfair and unethical. "You don't want to go down as the first company to enable and open the door for human-designed AI manipulation at scale," wrote leading AI researcher Clement Delangue on X, highlighting that Anthropic's decision to invisibly degrade answers related to AI development would be "the highest form of manipulation." Anthropic reacted quickly, updating its rules early Thursday morning to make such safeguards visible. Wired first reported Anthropic's reversal. Peter Wallich, a senior research manager at a AI safety research center called the Constellation Institute, said that the rocky rollout was less than ideal but sensible given the powerful technology. "It's clearly frustrating for security researchers and biologists to be kicked back to Opus for innocuous tasks, which is a real cost," Wallich told NBC News, emphasizing that he spoke in a personal capacity. "But this still seems to me like a reasonable trade-off. Regular people get the model earlier than they would otherwise." Wallich said that a more lax approach -- prioritizing earlier public access to Fable 5's full power without adequate protections -- would be a far worse scenario. "The opposite failure mode, shipping with safeguards that were too loose, would be irresponsible and could have led to irreversible harm," he said.

Anthropic Releases a Safer Version of Its 'Too Dangerous' Mythos AI

Back in April, Anthropic announced it had built a new AI model, called Mythos, that had an unprecedented ability to find gaps in cybersecurity systems. Citing the risk of the powerful new model falling into the wrong hands, the company initially withheld its public release, only slowly granting access to a small group of early testers through a project known as Project Glasswing. Today, Mythos finally made its first public debut -- albeit in a considerably nerfed fashion. Anthropic announced Tuesday that it's releasing a new model called Claude Fable 5, described in a company blog post as "a Mythos-class model made safe for general use." The new model's capabilities across tasks like coding, vision, tool use, legal analysis, and scientific research "exceed those of every model we've ever made generally available," the company wrote. (The implication there is that the most powerful version of Mythos is capable of more, but that it hasn't (yet) been deemed fit for public availability.) "The longer and more complex the task, the larger Fable 5's lead over our other models." The company has also launched another model called Mythos 5, which has the same capabilities as Fable 5, minus some of the safeguards. That model, however, is for the time being only available to the small cohort of organizations that are part of Project Glasswing, but Anthropic said access will be expanded soon. Safeguards Fable 5 comes with safeguards designed to prevent it from responding to questions on particularly sensitive subjects, like cybersecurity, biology, and chemistry. In those cases, it will automatically revert to an earlier Claude model, Opus 4.8, to avoid divulging information that could, say, make it easier for aspiring terrorists to cook up a biological weapon. Anthropic says it red-teamed the new generally available model, both internally and in collaboration with external testers, to root out security flaws that could lead to jailbreaks. (That is, manipulating an AI chatbot to ignore and bypass its built-in safety guardrails). Both Fable 5 and Mythos 5 are trained to play it safe, Anthropic said, which could lead to instances in which user queries that are in fact benign are erroneously flagged as dangerous by the model. "Because we have prioritized safety, we've deliberately tuned the safeguards to be cautious, and they are still stricter than would be ideal," the company wrote in the blog post. "We recognize that this will be frustrating to some users, and our aim is to reduce false positives as we update and refine the safeguards after launch." Opus 4.8 has already been annoying some users after Anthropic boosted the model's "honesty." Cost With a hefty price tag of $10 per million input tokens and $50 per million output tokens, Fable 5 and Mythos 5 cost twice as much as the standard version of Opus 4.8. At a time when many individual developers and organizations are starting to fret over the costs of using AI, that will likely push Anthropic's new model beyond the budgets of much of its userbase. But the powerful aura that's been built up around Mythos over the past couple of months also shouldn't be underestimated: Many people who have been hearing rumours about the fabled Mythos will likely be more than glad to foot the bill for token usage just to see if Fable lives up to the rumors.

Claude Fable 5: Anthropic releases a 'safe' version of Claude Mythos

Anthropic has released Claude Fable 5, a publicly available version of its powerful but previously restricted Mythos model -- complete with a new set of safety guardrails designed to keep its most dangerous capabilities out of the wrong hands. Along with this "safe for general use" model, Anthropic also released Claude Mythos 5, a version of Fable without the safety guardrails, to trusted testing partners. Earlier this year, Anthropic announced a limited launch of Claude Mythos, a new model with advanced cybersecurity capabilities that Anthropic deemed too dangerous to release. The company says Fable 5 is the most capable model it has ever made generally available, leading nearly all tested benchmarks across software engineering, knowledge work, vision, and scientific research. The more complex the task, Anthropic says, the wider Fable 5's edge over its previous models and competitors. Fable 5 shares the same underlying architecture as Claude Mythos 5 -- the restricted version shared with cybersecurity partners through Project Glasswing -- but ships with classifiers that intercept sensitive queries and route them to Claude Opus 4.8 instead. The restricted categories include cybersecurity, biology, and chemistry, as well as attempts to distill the model's capabilities for use in competing systems. Anthropic says fewer than five percent of sessions trigger a fallback, though it acknowledges the system is tuned conservatively and will occasionally flag benign requests. How to try Claude Fable 5 Fable 5 is available today across all Claude plans and via the API using the model string claude-fable-5. It is priced at $10 per million input tokens and $50 per million output tokens -- less than half the cost of Claude Mythos Preview. Subscription plan users get access at no extra cost through June 22, after which usage credits will be required. Benchmarks In agentic coding evaluations, Fable 5 outpaced GPT-5.5 and Claude Opus 4.8 by significant margins, according to Anthropic. The company's data shows that it even outperforms Claude Mythos on some key benchmarks. In a blog post, Anthropic wrote that fintech company Stripe, which had early access to Fable 5, reported that the model completed a full migration of a 50-million-line Ruby codebase in a single day. Anthropic estimated that this work would have taken a full engineering team more than two months. Fable 5, Mythos 5, and safety The safety story here is genuinely complicated. Anthropic spent months warning that Mythos-class models were too dangerous for general release. As recently as May, the company publicly acknowledged that adequate safeguards didn't yet exist, per prior Mashable reporting. Fable 5 is its answer to that problem, but the company's own disclosures suggest the solution is still a work in progress. An external bug bounty ran more than 1,000 hours of testing without producing a universal jailbreak -- but the UK AI Safety Institute made early inroads toward one in a brief initial window. Anthropic frames that as acceptable risk. Others may disagree. The Fable 5 system card states that the model has similar performance to Claude Opus 4.8 and other recent models on misaligned behaviors such as hallucination, dishonesty, and sycophancy.

Anthropic to reassess Claude Fable 5 AI development restrictions after backlash

Anthropic's AI restrictions caused concern among founders, researchers and developers following the release of the 'Mythos-like' model. It was just this week that Anthropic launched Claude Fable 5, an AI model that the company claimed has the potential to rival Mythos, albeit with significant security restrictions in place to prevent misuse. Already, however, the organisation is facing backlash from founders, researchers and developers who find the 'secretive' LLM policies to be deliberately limiting competitors and users in the development of alternative AI models. In a recent statement announcing the availability of Claude Fable 5, Anthropic explained that the model has barriers designed to block responses that stray into high-risk areas, notably, cybersecurity, chemistry and biology. Such Interactions, it stated, will instead be rerouted to the Opus 4.8, a less powerful model. This has reportedly drawn some criticism from researchers and developers in the AI space, as there is a concern that work derived this way is deliberately degraded in a manner that is invisible and secretive. Anthropic has since responded to the backlash, indicating plans to address the issue by making Claude 5's safety rails visible to all users and if the company suspects that a user is attempting to develop their own high-powered AI system, it will send a clear prompt informing them that the request is being refused based on policy, or that it is being rerouted to a less-capable model. It is currently against Anthropic's terms of service to use Claude technology as a means of training competing AI models. In a statement to Wired, a representative for Anthropic said, "We're changing Fable 5's safeguards for frontier LLM development to make them visible. We made the wrong trade-off and we apologise for not getting the balance right." Among the critics of Anthropic's decision-making was Dean Ball, a senior fellow at the Foundation for American Innovation and a former adviser to the White House on AI. In a post published on X, Ball described deliberately degrading ML research performance without informing the user as a "shockingly hostile and terrible look". He further explained its potential to "silently damage all sorts of work" and "raise the eyebrows of antitrust enforcers worldwide." Don't miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic's digest of need-to-know sci-tech news.

Anthropic sets AI performance records with new Mythos 5, Fable 5 frontier models

Anthropic sets AI performance records with new Mythos 5, Fable 5 frontier models Anthropic PBC today introduced Claude Mythos 5 and Claude Fable 5, two large language models that it says outperform the competition across a wide range of benchmarks. The LLMs are derived from the Claude Mythos Preview algorithm that the company debuted in April. The model made headlines for its ability to find highly complicated cybersecurity vulnerabilities. According to Anthropic, Mythos 5 and Fable 5 both surpass its performance. Fable 5 is broadly available to the public. It blocks prompts that relate to high-risk use cases such as cybersecurity vulnerability discovery. Mythos 5, by contrast, includes more relaxed guardrails and will only be available to a limited number of organizations. Anthropic PBC will work with the US government to manage access to the model. The company says that Mythos 5 is the first model in the Claude series with the ability to "consistently produce novel, compelling scientific hypotheses." The company asked the LLM to propose explanations for several molecular biology phenomena that scientists don't yet fully understand. According to the company, several of Mythos 5's suggestions were so promising that its researchers decided to launch a series of experiments aimed at verifying them. One of the model's hypotheses has already been corroborated in a lab. According to Anthropic, the model discovered new information about one of the proteins that make up the E. coli bacterium. Another internal test saw Mythos 5 discover 14 protein targets, biological building blocks that can potentially be used to make medicines. Anthropic claims that nine of the targets proved to be "strong candidates for drug design." Furthermore, Mythos 5 sped up some of the tasks involved in the protein discovery process by a factor of 10. Fable 5, Anthropic's other new LLM , blocks requests that relate to cybersecurity, biology and chemistry. The built-in guardrails reroute such prompts to Claude Opus 4.8, a less advanced model that Anthropic launched in May. Both Fable 5 and Mythos 5 have set a record on the SWE-Bench Pro programming benchmark with a 80.3% score. Stripe Inc., an early adopter of the former model, used it to modernize an internal software repository with 50 million lines of code. Anthropic says that Fable 5 enabled the company to complete the task in a day instead of the two months it would have otherwise required. Fable 5 and Mythos 5 also provide significant performance gains across several non-technical use cases. They scored 7.3% higher than Opus 4.8 on the GDP.pdf benchmark, which comprises document review tasks. Additionally, the models set a record on a second benchmark that measures LLMs' ability to automate legal tasks. Anthropic's new models are priced at $10 per million input tokens and $50 per million output. That's less than half what the company charged for Mythos preview. Furthermore, Mythos 5 and Fable 5 use fewer prompts to complete tasks.

Anthropic releases a version of its vaunted Mythos model to developers

The AI startup said Claude Fable 5 is more capable than any other model it has released to the public, showing outstanding performance in software engineering, knowledge work, and vision tasks, among other areas. The model is superior to earlier iterations in performing longer, more complex tasks. The analytics company Hex said that Fable 5 was the first model to break 90% on its core benchmark of complex, long-running analytical tasks-a 10-point jump over earlier Opus models. Claude Fable 5 scored 80.3% on the SWE-Bench-Pro, which measures AI models' agentic coding skill, compared to (OpenAI) GPT-5.5's score of 58.6%, and (Google Deepmind) Gemini 3.1 Pro's score of 54.2%. The testers said Fable showed "strong judgment and attention to nuance." Anthropic released the original Mythos Preview model in April, but only to a select set of defensive cybersecurity pros and overseers of critical infrastructure. Mythos, the company said, had proved to be very good at finding and exploiting security weaknesses in commercial software.

Anthropic accused of 'secret sabotage' as Claude Fable 5 silently limits capabilities for AI researchers and developers | Fortune

When Anthropic made its first Mythos-tier model available to the general public yesterday, called Claude Fable 5, Fortune reported it was a "considerable step" for the lab, coming just over a week after the company confidentially filed for IPO paperwork. It had initially deemed Mythos-class models too dangerous to release, citing their significantly enhanced ability to identify software vulnerabilities, but said it was now confident new guardrails in Claude Fable 5 are enough to ensure these dangerous skills don't fall into the wrong hands. Just hours after the model's release, however, major backlash from AI researchers, developers and policy experts began brewing on social media. The pushback centered around a paragraph buried in Claude Fable 5's 319-page system card -- a document that offers detailed safety disclosures -- which revealed that Fable would quietly downgrade its own responses when it detected requests related to cutting-edge AI development work, such as building the infrastructure used to train large AI models. In practice, that means a user could ask Fable for help, receive a deliberately weakened answer, but not know the model was holding anything back. Critics made it clear they felt this undermined a basic expectation that a tool would either do what it was asked or tell the user it wouldn't. Unlike Fable's other restrictions, such as around cybersecurity and biology, which openly redirect users to a less powerful model with a visible notification, the system card emphasized that this is "not visible to the user." The model still responds, but uses "interventions to limit Claude's effectiveness" without telling the user it's doing so. Anthropic estimated the restrictions would affect roughly 0.03% of traffic. But it also defended its effort by saying "enforcing this restriction through our safeguards avoids accelerating the actors most willing to violate these terms." Pushback from AI community A wide swath of the AI community pushed back sharply -- including open-source researchers critical of Anthropic's closed policies, as well as AI safety experts who typically align with Anthropic. "To have my access to the cutting edge models for my work rug pulled in an under the table fashion is appalling," wrote Nathan Lambert, an open model researcher who most recently led work at AI2. "To me this paints Anthropic clearly as anti-science, and therefore anti-progress and anti-safety." Dean Ball, a senior fellow at the Foundation for American Innovation who previously served as senior policy advisor at the White House Office of Science and Technology Policy, wrote that Anthropic's "secret sabotage" safety policy "massively and profoundly raises the status of the argument that AI safety has been hype to justify monopolistic behavior by labs." And Jeremy Howard, head of nonprofit research group Fast AI, wrote that "Anthropic has chosen the opposite of the safe path: they are allowing themselves, the current top lab, to use their top model for frontier AI research. They've said they'll sabotage others who try. This means the AI frontier advances, & power imbalance increases." Even former Anthropic employees joined in. Behnam Neyshabur, who previously co-led Anthropic's effort to develop an AI scientist, posted on X saying: "Working on AI for cancer? Sorry, I can't help you. Working on AI for Alzheimer's Disease? Sorry, I'm becoming a bit dumb when it comes to the AI part of it." In another post, he added: "I've argued for the last eight months that this was the direction things were heading. In my view, concentrating these capabilities fundamentally slows scientific and technological progress and is net negative for humanity." Not all prominent AI voices weighed in with criticism, however. Ethan Mollick, an associate professor at Wharton studying AI, innovation and entrepreneurship, did not focus on the restrictions, writing in a blog post that Claude Fable 5 "outperformed basically every other public model I have used by a considerable margin." Former OpenAI cofounder and Tesla AI director Andrej Karpathy, who announced he had joined Anthropic last month, called Claude Fable 5 a "super exciting release" on X and said it is a "major-version-bump-deserving step change forward." He did, however, point out that the model "still has quirks that people will run into and the safeguards are configured to be a little too trigger-happy for launch, which can hopefully be tuned over time." Anthropic says it wants to make models accessible and safe Before the release, Anthropic seemed to gird itself for backlash, though it did not specifically address potential blowback regarding the research restrictions. In an interview with Fortune yesterday, Dianne Na Penn, Anthropic's head of product management, research, and labs, said that the new model was able to produce frontier performance that was 10-20 points more than its previous model, Opus 4.8 or other frontier models. "I think generally being able to do that at the same time having the right guardrails in place to make it accessible, and generally in a safe manner, I think that's probably the main thing that I want folks to take away," she said. "We're raising the bar on the intelligence of the models, and at the same time, we are pushing the frontier in a safe manner." She added that Anthropic recognized that some benign requests would initially be blocked. "We're working actively on making those safeguards improvements post-launch, but we wanted to make the model accessible generally in a safe manner as soon as we could." Anthropic did not respond to Fortune's request for comment.

Anthropic Rolls Out Claude Mythos 5 AI Model -- Along With the Safer Fable 5 for the Public

Mythos drew scrutiny after tests showed it could autonomously discover vulnerabilities and complete complex cyberattack simulations. Anthropic on Tuesday announced the official launch of Claude Mythos 5, a restricted-access AI model the company says has unprecedented cybersecurity capabilities, alongside Claude Fable 5, a version designed for broader public use. In a blog post, Anthropic said Mythos 5 will initially be available only to approved cybersecurity organizations, critical infrastructure operators, government partners, and selected life sciences researchers. The company says the Mythos-class AI models "have reached a threshold where they present significant risks." "In April, we began Project Glasswing, releasing the first Mythos-class model (Claude Mythos Preview) to only a limited group of cyber defenders and critical software infrastructure providers," Anthropic wrote. "When we did so, we stated that we hoped to eventually release Mythos-level capabilities to all our users, so long as we had developed new safeguards that were strong enough to reliably prevent misuse." The launch follows months of concern over Mythos' ability to identify software vulnerabilities and conduct complex cybersecurity operations after news of the AI leaked online in March. Earlier this year, the U.K.'s AI Security Institute reported that a preview version became the first AI model to complete a 32-step corporate network intrusion exercise without human assistance. In April, Mozilla reported that a preview version of Mythos discovered over 271 vulnerabilities in the Firefox browser. Until now, Anthropic has restricted access to Mythos through its Project Glasswing program while researchers evaluated the risks. In a follow-up post on X, Anthropic said it would continue to expand access to Mythos "through a broader trusted access program," for cybersecurity work and biomedical research. The launch of Mythos also comes as tensions between Anthropic and President Donald Trump's administration appear to be easing after a public feud over the use of Claude for military operations. "In consultation with the U.S. government, we plan to steadily expand access to Claude Mythos 5, continuing our periodic addition of new partners, as well as pursuing a trusted access program that allows cybersecurity organizations to apply in a more systematic manner," Anthropic wrote, adding that users of Mythos Preview can upgrade to Mythos 5. Earlier this month, press reports said that Anthropic had embedded engineers at the National Security Agency to help deploy its Mythos AI model for cybersecurity operations, reportedly targeting China and Iran. Alongside Mythos 5, Anthropic released Claude Fable 5, a Mythos-class AI model that uses the same core design of Mythos 5 but with additional safeguards for the general public. "Fable 5's capabilities exceed those of any model we've ever made generally available," Anthropic wrote. "It is state-of-the-art on nearly all tested benchmarks of AI capability, showing exceptional performance in software engineering, knowledge work, vision, scientific research, and many other areas." According to Anthropic, Mythos 5 and Fable 5 achieved top scores on several software engineering, cybersecurity, biology, and reasoning benchmarks. The company reported a 78% score on ExploitBench, which measures vulnerability discovery and exploitation, and an 88% score on the Terminal-Bench coding benchmark. Anthropic said the results show the models can handle longer and more complex tasks than previous Claude systems. While Anthropic said Fable 5 performs especially well on lengthy, complex tasks, in order to prevent misuse, requests involving cybersecurity, biology, chemistry, or AI model replication are handled by Claude Opus 4.8 -- a safeguard the company said is triggered in fewer than 5% of sessions. "Releasing a model this capable comes with risks," Anthropic wrote. "Without safeguards, Fable 5's capabilities in areas like cybersecurity could be misused to cause serious damage." The launch comes as AI developers face growing questions about how to release increasingly capable systems. Anthropic has argued that Mythos could help defenders find and fix vulnerabilities before attackers exploit them. At the same time, critics, including OpenAI CEO Sam Altman, have accused the company of using "fear-based marketing" to sell its AI models.

Anthropic launches Claude Fable 5 ahead of $965 billion IPO

Anthropic has launched Claude Fable 5, its first publicly available model in the Mythos class, which features a safeguard system that automatically switches to a safer model in three high-risk areas. The Mythos class represents Anthropic's highest model tier, sharing a base model with the recently unlocked Claude Mythos 5. Fable serves as a mid-tier model with restricted access to safety-critical functions. Previously, access to Mythos systems was limited to over 40 elite technology companies due to misuse risks, resulting from unauthorized access incidents that intensified security discussions. Claude Fable 5 replaces the earlier model, Claude Opus 4.8, becoming the most capable publicly accessible model. The pricing for Fable 5 is set at $10 per million input tokens and $50 per million output tokens, marking a twofold increase compared to Opus 4.8. In comparison to the Mythos Preview, its cost is less than half of the previous offering, broadening access for more developers. From June 22, 2026, Fable 5 will be available free for users on Pro, Max, Team, and Enterprise plans, with immediate access via API. Fable 5 has been generally accessible in GitHub Copilot since June 9, 2026, facilitating swift integration into development environments. The model incorporates a safeguard architecture that routes requests in high-risk areas -- cybersecurity, biology and chemistry, and model capability extraction -- to the older Claude Opus 4.8. According to Anthropic, this fallback mechanism engages in fewer than 5% of sessions. Tests did not reveal any universal jailbreaks after over 1,000 hours of external red teaming. All Mythos-class traffic is subject to a 30-day data retention rule and will not contribute to model training. Anthropic positions itself as the first provider to implement a formal structure for the controlled release of models deemed too dangerous for unrestricted public access. Claude Mythos 5, with some safeguards lifted, is available to selected partners such as AWS, Microsoft, Apple, and CrowdStrike, as well as approved bio-researchers. In benchmark comparisons, Fable 5 exhibits substantial improvements over Opus 4.8. It achieves 95.0% on software engineering benchmarks like SWE-bench Verified, compared to Opus's 88.6%. On SWE-bench Pro, Fable 5's lead increases to 80.3% versus 69.2%. Specialized tests reveal Fable 5 scored 29.3% on FrontierCode Diamond, compared to 13.4% for Opus 4.8. In practical applications, Stripe migrated over 50 million lines of Ruby code in one day utilizing Fable 5's capabilities. Additionally, the ExploitBench results indicate a significant performance gap in offensive capabilities, with Mythos 5 scoring 78.0%, in contrast to 40.0% for Opus 4.8. This performance gap highlights the importance of the fallback system integrated within Fable 5's public access model. On June 1, 2026, Anthropic confidentially filed a draft S-1 form with the SEC for an IPO, valuing the company at $965 billion after a $65 billion funding round. The revenue run rate was approximately $47 billion as of May 2026, anticipated to exceed $50 billion by the end of June, boosted by significant growth from a run rate of $4 billion in July 2025. Coinciding with this filing, OpenAI also confidentially submitted for a listing on June 8, 2026, indicating a potential maturation phase in the AI sector.

Anthropic releases state-of-the-art AI model Claude Fable 5, and it's safe to use

Anthropic has officially launched Claude Fable 5, a Mythos-class AI model that now available to the public, and to celebrate the launch Anthropic has also reset current and weekly session limits. The release marks a major milestone for the company, offering users unprecedented access to a powerful system that balances performance with safety. Anthropic explains in its blog and a series of videos that after it conducted its preview of the Mythos model and discovered thousands of cybersecurity vulnerabilities, which meant the Mythos model could be used for nefarious practices if used by the wrong hands. So, Anthropic created a new model based on the Mythos model, but implemented safeguards. Fable 5's capabilities surpass those of previous models, excelling in complex reasoning, natural language understanding, and multi-step problem solving. According to Anthropic, the model has been made safe for general use while retaining the advanced features of its Mythos sibling. The company also writes that if users begin asking questions in the following areas Fable will detect the line of questioning and prevent any misuse of its capabilities. Those areas are: cybersecurity, biology, chemistry, and distillation. This release positions Anthropic at the forefront of AI arms race. With Fable 5, the company is making a push to democratize access to high-level AI tools while maintaining ethical guardrails. This approach contrasts with the more restricted access of the Mythos series, which is typically reserved for enterprise and research use.

Crypto Users Wary as Anthropic's Claude Mythos Goes Live

Venture capitalist Simon Dedic said Anthropic's latest AI models drop the cost and skill needed to find crypto exploits to "basically zero." AI company Anthropic on Tuesday released the first public version of its powerful Claude Mythos model, called Fable 5, with some crypto users worried it could be used for malicious purposes, despite embedded guardrails. Anthropic said last month that its Mythos model uncovered more than 10,000 high or critical-severity vulnerabilities in "systemically important software," leading many to question if it should be publicly released. This was despite the company saying on Tuesday that Fable 5 was "made safe for general use," and has safeguards that reroute some topics, such as cybersecurity, to a different model, Claude Opus 4.8 "Releasing a model this capable comes with risks. Without safeguards, Fable 5's capabilities in areas like cybersecurity could be misused to cause serious damage," it said. Source: Claude The guardrails have done little to reassure crypto users, with AI increasingly being used to attack crypto platforms. In April, the value of crypto stolen in hacks hit $629.7 million, the highest since February 2025, which analysts linked to the use of the technology. Mythos release sparks warnings from crypto users Simon Dedic, founder of the venture firm Moonrock Capital, posted to X on Tuesday that with Fable 5, the "cost and skill required to find exploitable flaws in smart contracts is about to drop to basically zero." "For DeFi, this should be a massive wake-up call. Unaudited protocols will become sitting ducks. Known exploits will get replayed on forks around the clock. Even small projects will get targeted simply because trying costs next to nothing now," he added. Dedic repeated calls online, suggesting that crypto users should protect themselves from the model, including revoking wallet approvals, removing as much value from protocols as possible and moving crypto to fresh hardware wallets. Curve Finance co-founder Michael Egorov, however, said that the threat Claude Mythos posed to crypto was likely overblown as its success in finding bugs in other software might not translate to funding smart contract vulnerabilities in DeFi. In May, Anthropic said Claude Mythos found thousands of critical vulnerabilities in important software through Project Glasswing. For open-source projects, which are central to how crypto protocols are managed, Mythos found around 6,200 high or critical-severity vulnerabilities in more than 1,000 projects it investigated. Egorov argued that the software Mythos found vulnerabilities in had millions of lines of code, while smart contracts have a few thousand, "and both humans and 'usual' AI perfectly fit that code in context and can reason well about it." "I suspect we might not be having a wave of DeFi code hacks, but we may see a lot of things in OpSec [operational security] getting hacked (looking like multisig keys compromises) and supply chain attacks on frontend dependencies, and those are way less dangerous in true DeFi," he said. Meanwhile, Anthropic said a "small group" of cybersecurity and infrastructure providers would get access to Claude Mythos 5, the same model as Fable 5 but with safeguards lifted in some areas.

Anthropic releases Fable 5 model, built on the same tech that spooked the government

The Anthropic logo as seen at an event organized by the AI company. Andrej Sokolow / picture alliance via Getty Images file Anthropic released its latest model Tuesday afternoon, heralding the public's first access to the AI company's most powerful class of AI systems. The company says the model, termed Fable 5, is the first publicly available product in the same family as Anthropic's powerful Mythos models, which sent shockwaves through the cybersecurity world earlier this year for their superhuman ability to find and exploit cyber vulnerabilities. "Fable's capabilities exceed those of any model we've ever made generally available," the company said in a blog post announcing the model's release. "It is state-of-the-art on nearly all tested benchmarks of AI capability, showing exceptional performance in software engineering, knowledge work, vision, scientific research, and many other areas." According to the company, Fable 5 uses the same tier of technology as Mythos but is safe for use by the public because of safeguards and limits placed on the technology. Anthropic's Mythos Preview model, which lacks similar safeguards, was able to find thousands of critical and severe cyber vulnerabilities, including bugs and exploits in all major operating systems and web browsers. Many AI researchers worry that increasingly powerful AI systems could help bad actors carry out cyberattacks on banks, power grids or other critical infrastructure. Others hypothesize that increasingly intelligent AI systems could help terrorists design and deploy bioweapons. "The same queries that are beneficial in the hands of cybersecurity professionals and biology researchers could be dangerous if available to malicious actors," the company wrote in its blog. To address these potential threats, Anthropic has said Fable 5 is being deployed with guardrails that block many of its responses to queries regarding potentially dangerous topics. For those user prompts, Anthropic will steer answers to an earlier, less-powerful model called Opus 4.8 -- which was the highest-performing publicly available model until Tuesday. "Fable 5's capabilities in areas like cybersecurity, biology and chemistry are advanced enough that we're taking a deliberately conservative approach for these topics at launch," Anthropic wrote in reply to a question from NBC News. "To enable general availability of other Mythos level capabilities, we've decided to deploy safeguards that err on the side of caution, applying broad restrictions to these topics for now." Anthropic said the two-pronged approach, diverting sensitive questions from Fable 5 to the older Opus 4.8, would allow users to still obtain helpful answers to questions when Fable 5's capabilities could prove too dangerous. Anthropic also announced that the trusted partners who had previously been able to access Mythos Preview will now be able to access an upgraded Mythos model, called Mythos 5. The company had made Mythos Preview available to over 150 organizations around the world to help financial institutions, software companies and healthcare networks shore up weaknesses in their cyber defenses prior to a wider public release of a Mythos-class model. Anthropic said Mythos 5's capabilities shattered existing performance records across a range of other domains, including drug design and molecular biology. Anthropic said Mythos 5 is the company's "first model to consistently produce novel, compelling scientific hypotheses." Tuesday's model release comes one week after President Donald Trump signed a new executive order that aims to establish a new, voluntary mechanism for AI companies to share their systems with the government for safety testing before they are publicly released and to shore up the government's own cyber defenses.

Claude Fable 5 Is Here, But You Probably Won't Need It

If you've been using AI for a while, you probably already know about Claude. A lot of people do. But when most folks talk about AI, they say ChatGPT, the same way people say "Google it" when they mean search. Anthropic has never quite broken into that default. Claude Fable 5 might be its best shot yet. Anthropic launched Claude Fable 5 and Claude Mythos 5 yesterday, and by its own description, these are the most capable models the company has ever released, surpassing the entire Opus line. That's a big deal on its own. But the backstory makes it a bigger one. When Anthropic first unveiled Mythos back in April, the company said it wasn't safe for public hands. The model could autonomously find and chain zero-day exploits across major operating systems and browsers. Anthropic limited access to a small group of vetted partners through a program called Project Glasswing. Now, two months later, there's a version for everyone. So What Changed? Fable 5 and Mythos 5 are the same underlying model. The difference is the guardrails. With Fable 5, queries touching on sensitive cybersecurity, chemistry, or biology topics automatically fall back to Opus 4.8 instead. Anthropic says this triggers in less than 5% of sessions. Mythos 5 keeps those safeguards lifted and stays restricted to cyberdefenders and infrastructure providers working alongside the US government. Anthropic has Fable 5 on Pro, Max, Team, and Enterprise plans through June 22. After that it shifts to usage credits until capacity catches up. That being said, Claude Fable 5 probably won't feel all that different from what you've been using. If you're asking it to help write something, summarize an article, or answer a question, you're not really the target audience here. Fable 5 is built for people doing serious technical work, like security researchers and developers pushing up against what AI could previously handle. For everyone else, it's still a solid upgrade. Just not a dramatic one. Which brings it back to the bigger picture. Anthropic has been quietly making moves, from buying Super Bowl ads to landing deals with Microsoft, and now releasing the most powerful model it's ever made to the public. ChatGPT still has the name recognition. But the gap could be closing.

Anthropic Just Released a Powerful Mythos-Class Model to the Public -- With Some Key Safeguards

Anthropic has finally released a version of Claude Mythos, its most powerful model yet, but most people will experience it under a different name. The company just announced Claude Fable 5, its first commercially available Mythos-class model. The Dario Amodei-led company says that the new model's capabilities "exceed those of every model we've previously made generally available." The model is the first in Anthropic's new line of Claude 5 models. Anthropic has been preparing for this moment since April, when it first unveiled Claude Mythos, a line of extremely large, capable AI models. The company didn't make Claude Mythos available to the public initially in order to give cybersecurity professionals time and access to the tool in order to strengthen their defenses for a new era of hacking, as the model is prodigious at identifying and exploiting software bugs. Mythos has dominated conversation in AI, and has hastened conversations around regulating AI models. Now, Anthropic says that it has figured out a method to safely release a Mythos-level model to the public. The company is launching Fable 5 with safeguards that block some responses related to cybersecurity and biology (as models are getting better at building bioweapons), and will instead be routed to a response from Claude Opus 4.8. Anthropic says that its cybersecurity red team ran a program to identify potential methods for getting around these safeguards, but was unable to find a single workable jailbreak in over 1,000 hours of testing.

Claude just released a Mythos-level model, but you only have 10 days to try it

Outside of the office, Josh can be found digging into the latest video games, fantasy books, or tinkering with the newest features in Windows. * Anthropic has released Fable 5 -- a Mythos-level Claude model available for a limited time. * Tuned for cybersecurity and biology and features robust safeguards to drop it to cheaper models if triggered. * Subscribers can use Fable 5 until June 22; expect high cost and token burn. If you've been itching to get your hands on Claude's Mythos model -- the company's top-tier model designed for deeper coding and even spotting bugs and cybersecurity issues in other apps and systems -- then now might be your only chance, at least for a little while. That's because Anthropic has hot-dropped a new model for Claude subscribers to test out called Fable 5, and it's only out for a limited time. Anthropic just doubled limits for Claude plans, and it could transform how you work Perfect for anyone who uses Claude at work. Posts 1 By Sara Heritage A taste of what Claude can really do It's Mythos-level, supposedly According to Anthropic's announcement post about Fable 5, the new model appears to be at least on some level similar to Mythos, as it is actually labeled next to Mythos 5 on a performance graph. The company says it is built to tackle your longest-running and most complex usages. Additionally, based on the performance graph seen in the announcement, it even performs better than the Mythos Preview in some cases. Now, none of this is to downplay how impressive Claude can be, especially if you utilize it correctly. For instance, using Claude Cowork to take over repetitive tasks can be game-changing. If you haven't been paying attention to the conversation surrounding Claude Mythos, the model is supposedly so dangerously good at what it does, that Anthropic is afraid to release it to the general public. Instead, the company has been testing it behind closed doors, as part of a larger cybersecurity push to bring more AI into that particular sector of things. That, itself, is terrifying enough. However, Fable 5 appears to be a more consumer-focused version of the Mythos model. Though, exactly how much has changed, or how much Anthropic has limited it, is hard to say without a direct comparison. However, the model is clearly more geared towards cybersecurity and biology industries, as Anthropic even notes in the announcement that there are "robust safeguards for cybersecurity and biology" that will kick the AI model to a less expensive model should any of the safeguards be hit. This should help cut down on usage, which is twice the cost of Claude Opus, according to a note on the Claude website and app. A limited time to try it After June 22, you'll need to dish out extra credits for access Which brings us to the biggest crux of this announcement. Despite the supposed advancements that Anthropic is flaunting on its website for Fable 5, access to the new model is going to be limited. Subscribers to Claude can try it out right now, but not only is it expensive (as noted above), but it will stop being available on June 22. That means you have just over 10 days to toy around with it before Anthropic takes it off the list and locks it behind additional credit purchases. We also don't really know how well it will stand up against Mythos in day-to-day usage, as nobody in the general public has had a chance to really toy with Mythos yet, thanks to its extremely limited access. And that doesn't seem to be changing, as Mythos 5 will continue to only be available to Glasswing members (which is the cybersecurity alliance that Anthropic set up with several other major tech companies). But it is pretty powerful, as Anthropic has already shown off that it can beat Pokémon with just visual input. Claude Developer Anthropic PBC Price model Free, subscription available Claude is an advanced artificial intelligence assistant developed by Anthropic. Built on Constitutional AI principles, it excels at complex reasoning, sophisticated writing, and professional-grade coding assistance. See at App Store See at Google Play Store See at Claude Expand Collapse If you're Claude Pro subscriber, then this is definitely one of those Claude features you need to take advantage of, especially if you like messing around with the latest and greatest AI models, then Claude has a new one to try. Just don't be surprised if you run out of tokens within a couple of prompts.

Anthropic rolls out 'Mythos-like' AI model Claude Fable 5

The model reportedly has safeguards built within it to prevent misuse by those who might exploit the technology. Just two months after rolling out Mythos to a limited pool of high-level users, Anthropic has announced the release of Claude Fable 5, an AI model similar to Mythos, but with significant safeguards and blocks to prevent deliberate misuse and security breaches. Unlike Mythos, which is currently only available to a select number of organisations and institutions due to major concerns about securing critical infrastructure, Claude Fable 5 will be made available to enterprise customers and paid subscribers. The model has built in barriers that block responses in high-risk areas such as cybersecurity, chemistry and biology, but it reportedly shows strong capabilities in software engineering, knowledge work, vision, scientific research and similar areas. In a statement, Anthropic explained that over the course of the past few months, the organisation has worked to improve safeguards that would make Claude Fable 5 "robust enough for a general release". Adding that in prioritising safety some measures are "stricter than would be ideal", as some benign requests may be classified as risky. However, there are plans to further refine the regulations. Anthropic has also announced an updated version of the Mythos model, Claude Mythos 5, which reportedly is similar to Fable 5, however, with the cyber safeguards lifted. The organisation said, "In consultation with the US government, we plan to steadily expand access to Claude Mythos 5, continuing our periodic addition of new partners, as well as pursuing a trusted access programme that allows cybersecurity organisations to apply in a more systematic manner." In early June, Anthropic unveiled plans for a historic initial public offering (IPO) that would take the company's valuation soaring above $1trn. The proposal came less than a week after the company overtook OpenAI's valuation with a $65bn Series H funding round that valued it at $965bn. Don't miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic's digest of need-to-know sci-tech news.

Anthropic releases its first Mythos-class model to the public | Fortune

Now, Anthropic says it is confident new guardrails are enough to ensure these dangerous skills don't fall into the wrong hands. "The reason why we're releasing Fable Five now is very much due to us feeling more confident with our safety guardrails in place," Dianne Penn, Anthropic's head of product management, research and labs, told Fortune. The company said that responses in specific high-risk areas, such as biology and cybersecurity, will be blocked and will, in most cases, be answered by Claude Opus 4.8, Anthropic's less powerful model, released earlier this year. The fear is that giving users free access to Mythos-level intelligence could enable bad actors to carry out more advanced cyberattacks or develop bioweapons more easily. The company also said it had extensively red-teamed its classifiers (machine learning algorithms) to test their robustness against jailbreaks. Anthropic says the new model capabilities "exceed those of every model we've previously made generally available," and demonstrate exceptional performance in areas such as coding, knowledge work, and vision. Penn said that Fable Five is particularly strong at what is known as "long-horizon memory management" -- an AI model's ability to keep track of what it's doing over a long, complex task. Where previous models sometimes "lose the thread" partway through complex, long tasks, the new model is much better at remembering what it's doing over long stretches, she said. Penn also said the lab has seen improvements in self‑verification and instruction following: the model is more deliberate about checking its own work, validating assumptions, and course‑correcting before producing a final answer. These changes mean higher overall quality across both coding and non‑coding tasks, she added. "We think there are some use cases where it's a replacement [for knowledge work], whether it's coding or others, but we also think that it's really expanding the use cases," Penn said. "Now you could use a model like Fable Five to run very long-horizon projects, potentially overnight ... and that might include things like being able to review your whole code base for improvements." "We recommend that customers give their most challenging work to Fable Five," she added. The lab said that early data shows at least 95% of Fable sessions run entirely on Fable's own responses, rather than being routed back to Opus. However, Anthropic has faced backlash from some users in the past who claim that the company's safety filters occasionally "over-block" benign requests or issue false refusals. Penn acknowledged the guardrails might not be perfect the first time around. "We recognize that there might be some benign requests that end up being blocked initially," she said. "We're working actively on making those safeguards improvements post-launch, but we wanted to make the model accessible generally in a safe manner as soon as we could." Fable 5 and Mythos 5 are being offered at $10 per million input tokens and $50 per million output tokens -- twice the price of the standard version of the next most advanced model, Opus 4.8.

Fable 5: AI Heroics Battle with High Costs, Token Needs

Anthropic's latest AI model Fable 5 has the potential to transform IT operations with its superior coding abilities, but concerns over high compute costs and token consumption may limit widespread adoption, experts said. "Fable's biggest differentiator is its agentic coding capability, which means it can accomplish complex tasks autonomously and find multiple paths to reach a goal," said Srikanth Velamakanni, chief executive at AI services firm Fractal Analytics and chairperson of Nasscom. "That makes it one of the most powerful software engineering models we have seen so far, with the potential to significantly transform enterprise IT operations, provided it is used responsibly." Still, models like Fable can suffer from "jagged intelligence," which means they are able to perform advanced tasks such as coding or passing medical exams but struggle with basic logic, spelling or simple counting, he said. Anthropic has priced Fable 30% cheaper than rival OpenAI's GPT 5.5 Pro model but it's currently slow and token hungry. Researchers suggest Fable is an enterprise grade model and not suitable for everyday common tasks. "Think of it like a bazooka -- it has a lot of power and consumes a lot of ammunition. In this case, ammunition (tokens) is very expensive," Velamakanni said. The latest release, Claude Fable 5, is a Mythos-class AI model with built-in safety guardrails. Anthropic has thus far kept Mythos under private preview. Experts said they are essentially the same model. They added that safeguards and levels of access vary. Early users have described Fable as the most capable coding model yet. It can autonomously assign complex tasks to agents and execute them as well. The model has beaten all industry performance benchmarks in biology, legal, spatial reasoning, agentic coding, cybersecurity and health, data showed. For instance, Fable migrated 50 million lines of code for Stripe in one day, a task that would typically require at least two months of human effort by a full team. It designed drugs 10x faster than humans and could even draft an S-1 filing for the SpaceX IPO from scratch that was nearly identical to the actual document. It could clone 3D games like Pokemon and Minecraft with a few simple prompts. The model's capabilities also pose a threat to India's $310 billion IT industry as agents automate grunt work. "Fable appears to be strong at modernization and maintenance activities when properly supported by context and tools," said Gary Olliffe, distinguished vice president, analyst, at Gartner. "This is likely to pose a challenge to the IT services industry in general, as expectations shift." However, cost concerns may keep the model's demand elastic, he said. Although the model's capability is notable, AI must be effectively embedded in "enterprise workflows, integrated with data, governed for risk and compliance, and scaled through robust operating controls," said Nitin Bhatt, partner and technology sector leader at EY India. "This reinforces the role of IT services firms in orchestrating and delivering enterprise-wide intelligence at scale." Fable comes with guardrails that would prevent its use in cyber exploits, said Vinayak Godse, chief executive, Data Security Council of India, part of industry grouping Nasscom. "Models like Mythos and Fable are raising the bar for what AI can accomplish, but they are also forcing organisations to rethink cybersecurity," he said. "Organisations are facing growing challenges in securing their codebases, digital assets and third-party software supply chains. I would say, from a cyber perspective, we need to be paranoid and worried, but not scared." Fable 5 is Anthropic's first Mythos class model that's safe for general use, said Varun Mishra, principal analyst at Counterpoint Research. "It is specifically designed to execute multi-hour autonomous tasks without human intervention straight out of the box, which is its key differentiator. However, its premium price tag and long horizon workflows would also mean a rising token bills for companies," he said. (This story has not been edited by economictimes.com and is auto-generated from a syndicated feed we subscribe to.)

Anthropic's latest AIs are making some customers uneasy

Catching a glimpse of Dario Amodei these days is like finding a rare butterfly. The chief executive officer of Anthropic PBC was scheduled to meet with 50 fellow bosses from some of Europe's largest companies in Oxfordshire recently, an exclusive convocation by his firm at a Jacobean-style mansion that included former British Prime Minister Rishi Sunak, now an Anthropic adviser. Amodei had been scheduled to travel to another part of Europe afterwards, but his trip was cut short to just one or two days in the U.K. Why is he so hard to pin down? Because of Mythos, the artificial-intelligence model he teased last month as too dangerous to launch. Anthropic staff have access to the technology and say it is more human than its forerunners. But it could also be weaponized for cyberattacks, so the company has let a few dozen organizations test and preview it. The model's power and potential will help Amodei's efforts to repair relations with the Trump administration and he's eager too for Mythos to finally reach members of the public. That requires a juggling act that, for now, Amodei is managing to pull off.

The Hidden Trade-Offs in Anthropic's New Mythos 5 and Claude Fable 5 Release

Anthropic has officially launched two advanced AI models, Mythos 5 and Claude Fable 5, each catering to distinct user needs with a focus on safety and performance. Mythos 5, part of the exclusive "Project Glass Wing" initiative, is designed for high-stakes applications where precision is paramount, making it accessible only to select enterprise clients. In contrast, Claude Fable 5 offers a more versatile solution aimed at a broader audience, balancing advanced functionality with wider accessibility. Sam Witteveen explores how these models differ in their intended use cases and the implications of their pricing and safeguards. Dive into this update to explore three key insights: how Mythos 5's exclusivity impacts its adoption in specialized industries, the practical trade-offs of Claude Fable 5's broader accessibility and the challenges posed by the models' stringent safety protocols. You'll also gain a clearer understanding of the cost considerations tied to these models and how they might influence smaller organizations or individual developers. This breakdown provides a thoughtful look at the opportunities and constraints these AI advancements bring to the table. Key Features and Differentiation Mythos 5 and Claude Fable 5 reflect Anthropic's dual approach to exclusivity and accessibility, offering tailored solutions for different user bases: * Mythos 5: A premium AI model designed for high-stakes applications where precision and safety are critical. Its exclusivity ensures it is reserved for enterprise clients with specialized needs. * Claude Fable 5: A versatile, general-purpose model intended for wider adoption. It balances advanced functionality with broader accessibility, making it suitable for a range of professional and creative tasks. Both models incorporate robust safety protocols, underscoring Anthropic's commitment to responsible AI development. This differentiation allows users to select a model that aligns with their specific requirements, whether for specialized industries or general use. Performance Benchmarks: Strengths and Limitations The performance of Mythos 5 and Claude Fable 5 has been rigorously tested, revealing significant advancements over earlier models like Opus 4.8 and GPT 5.5. These improvements are particularly evident in specialized domains: * Legal Analysis and Biology: Both models excel in handling complex, domain-specific tasks, demonstrating a high level of accuracy and contextual understanding in these fields. * General Reasoning and Creativity: Incremental improvements have been observed, showcasing enhanced problem-solving capabilities and creative outputs. However, there is still room for refinement in broader, less specialized applications. These benchmarks highlight the models' potential, but their true effectiveness will depend on real-world testing across diverse scenarios. Users will need to assess how well these models perform in practical, day-to-day applications. Unlock more potential in Claude Mythos by reading previous articles we have written. Cost and Accessibility: A Balancing Act The pricing structure of Mythos 5 and Claude Fable 5 reflects their advanced features but raises important questions about affordability and accessibility: * Mythos 5: Positioned as a high-cost option, it is tailored for enterprise clients with specialized needs and significant budgets. Its exclusivity may limit its adoption to a niche market. * Claude Fable 5: While more affordable, it remains a premium choice. Initially available under Pro and Max subscription plans, its pricing will transition to an API-based model after June 22, potentially impacting smaller organizations and individual developers. This pricing strategy highlights the challenge of balancing innovation with accessibility. Smaller organizations may find the cost prohibitive, raising concerns about the inclusivity of these advanced tools. Safety Measures: Enhanced Protections with Trade-offs Anthropic has prioritized safety in the development of Mythos 5 and Claude Fable 5, implementing stringent safeguards to prevent misuse and ensure ethical deployment: * Advanced monitoring systems are in place to detect and block attempts to bypass safety protocols, making sure the models operate within predefined ethical boundaries. * Restrictions on generating responses involving long chains of reasoning have been introduced to reduce the risk of harmful or unintended outputs. While this enhances security, it may limit the models' flexibility in certain scenarios. These measures reflect Anthropic's focus on responsible AI usage. However, users may need to adapt to these limitations, particularly in applications requiring greater flexibility or creative freedom. Data Retention Policy: Security vs Privacy A new 30-day data retention policy has been implemented for Mythos-class models, aimed at improving security and enhancing model performance. This policy, while beneficial in some respects, raises concerns about data privacy: * Users must carefully weigh the benefits of enhanced security against the potential risks associated with storing sensitive data for extended periods. * Organizations should evaluate how this policy aligns with their existing data governance practices, particularly in industries with strict privacy regulations. This policy underscores the ongoing tension between advancing AI capabilities and maintaining user trust through robust privacy protections. Use Cases and Practical Challenges Both Mythos 5 and Claude Fable 5 are designed for tasks requiring advanced problem-solving and strategic thinking. Their practical utility, however, depends on user-specific needs and the ability to integrate these models into existing workflows: * Claude Fable 5: Well-suited for generating detailed plans and creative outputs, but users may need to pair it with other tools for execution and implementation tasks. * Cost Considerations: The high price point of both models may deter smaller organizations or individual users, raising questions about their cost-effectiveness in less specialized applications. These challenges highlight the importance of aligning the models' capabilities with the practical requirements of their intended user base. Future Outlook: Questions and Adaptations As Mythos 5 and Claude Fable 5 enter the market, several key questions remain about their long-term impact and adaptability: * Will Claude Fable 5 remain accessible under its current subscription tiers, or will pricing adjustments further limit its availability? * How will users adapt to the stricter safeguards and the new data retention policy, particularly in industries with unique compliance requirements? * What updates and refinements will Anthropic introduce based on user feedback and real-world performance data? The answers to these questions will shape the future trajectory of these models. Ongoing testing, user feedback and iterative improvements will be critical to making sure their success and relevance in an evolving AI landscape. Media Credit: Sam Witteveen Disclosure: Some of our articles include affiliate links. If you buy something through one of these links, Geeky Gadgets may earn an affiliate commission. Learn about our Disclosure Policy.

Anthropic Rolls Out Fable 5, Mythos 5 -- One Comes With Limits

Anthropic launched Claude Fable 5, its most capable consumer-facing AI model to date, but with a catch: some sensitive cybersecurity requests are automatically downgraded to a less powerful model. The company said Fable 5 excels at coding, research and other knowledge-intensive tasks, particularly over long and complex workloads. To curb potential abuse, Anthropic built in guardrails that redirect certain high-risk queries to Claude Opus 4.8. The filters activate in fewer than 5% of sessions on average, according to the company. Mythos 5 uses the same underlying model as Fable 5 but drops some of the restrictions -- for a select group of cyber defenders, critical infrastructure operators and other trusted partners. The company said Mythos 5 will first be delivered through Project Glasswing in partnership with the U.S. government as an upgrade from Claude Mythos Preview, with broader access planned through a trusted program. Anthropic set pricing for both Fable 5 and Mythos 5 at $10 per million input tokens and $50 per million output tokens. The company said that rate is less than half the price of Claude Mythos Preview. Pokémon By Day, Protein Design By Night The models can sustain longer autonomous work than prior Claude versions and can stay on track across very long contexts while using notes to refine outputs. The company described tests in games such as "Pokémon FireRed" and "Slay the Spire" as demonstrations of those capabilities. On the life sciences side, internal protein design specialists used Mythos 5 to speed up parts of drug design work by roughly tenfold, and that several targets produced promising candidates now under review. The company also noted that Mythos 5 has generated original hypotheses in molecular biology and completed a week-plus genomics project that produced a custom model it said exceeded a recently published approach despite being far smaller. An automated alignment assessment found low levels of problematic behavior for Mythos 5, similar to Opus 4.8, and it expects comparable results for Fable 5 given the shared core model. Anthropic previously shared an update on a security-focused collaboration, Project Glasswing. Its artificial intelligence-assisted security testing effort had already uncovered "more than 10,000 high-or critical-severity vulnerabilities" across widely used software systems. Project Glasswing previously included roughly 50 partner organizations and later expanded access to nearly 150 organizations. Anthropic IPO Plans Last week, Anthropic confidentially filed a draft Form S-1 with the U.S. Securities and Exchange Commission, giving the artificial intelligence company the option to go public after the SEC completes its review. Anthropic said it has not yet determined how many shares it would sell or what price it would seek in an offering. Last month, Anthropic overtook OpenAI as the world's most valuable startup after raising $65 billion in Series H, valuing the company at $965 billion. Altimeter Capital, Dragoneer, Greenoaks and Sequoia Capital led the funding round. If Anthropic goes public at a $1 trillion valuation, it would instantly rank among the most valuable companies globally and could become the second- or third-largest IPO in history, trailing SpaceX and Saudi Aramco. Market News and Data brought to you by Benzinga APIs To add Benzinga News as your preferred source on Google, click here.

Claude Fable 5 & Mythos 5: Key highlights from Anthropic's latest launch

Anthropic has launched Claude Fable 5, its most capable publicly available AI model, excelling in complex tasks and benchmarks. Alongside it, Mythos 5, a restricted version with fewer safeguards, is available for trusted cybersecurity and life sciences users, aiming to accelerate research and discovery. Anthropic on Wednesday launched Claude Fable 5, a Mythos-class model approved for general use, and Claude Mythos 5, a restricted-access version with fewer safeguards for trusted cybersecurity and life sciences users. Here are some key highlights about these models. What are Fable 5 and Mythos 5? Fable 5 is Anthropic's most capable publicly available model, which leads key benchmarks across software engineering, knowledge work, vision (ability to understand images), scientific research, memory, and long-context tasks. Its advantage grows on longer and more complex workloads. Mythos 5 is the same underlying model as Fable 5 but with certain safeguards removed. Anthropic also highlighted major gains in analytical reasoning, drug design, molecular biology hypothesis generation, and genomics research. The company claims Mythos 5 can accelerate parts of drug discovery, generate novel scientific hypotheses, and conduct extended autonomous research. Safety testing found Mythos 5's alignment performance (measures how well its behaviour matches instructions) comparable to Opus 4.8, with low levels of deceptive or misuse-related behaviour. Anthropic also introduced stronger safeguards against jailbreaks and misuse, claiming significantly improved resistance compared with previous models. To mitigate misuse risks, particularly in cybersecurity and biological research, Fable 5 uses new safety classifiers. Requests related to cybersecurity, biology, chemistry, or model distillation (where a large model transfers its knowledge to a smaller one) are automatically routed to Claude Opus 4.8. Rather than issuing outright refusals, the Opus model responds to sensitive queries using more refined safety filters to distinguish legitimate research from misuse. These safeguards affect fewer than 5% of sessions on average. For Mythos-class models, the company stated it will retain business customer data for 30 days for safety monitoring purposes, while stating it will not use the data for model training. What about cost and access? Both Fable 5 and Mythos 5 are priced at $10 per million input tokens and $50 per million output tokens. Fable 5 is available immediately through the Claude API and on eligible Claude subscription plans during an initial rollout period. Meanwhile, Mythos 5 is currently available only to select cyber defenders and infrastructure providers through Project Glasswing, with broader trusted-access programs planned for cybersecurity and life sciences researchers. The launches come just weeks after India gained access to Claude Mythos Preview, Anthropic's advanced cybersecurity-focussed AI model, as part of an expansion of its Project Glasswing initiative. Why now? The launch of these models marks a major step in Anthropic's long-term effort to make Mythos-class AI capabilities more widely accessible. "Our goal is to safely open up access to vetted partners for Mythos 5's dual-use capabilities, to stay up to date on our plans for trusted access," the company wrote in its blog post. Dual-use refers to technology that has both civil or commercial and military or malicious use, like drones or satellites. The launch comes shortly after the company confidentially filed for an initial public offering (IPO) with the US Securities and Exchange Commission. Anthropic has seen strong business momentum, reporting a revenue run rate of $47 billion in May, up sharply from about $10 billion a year earlier. The company also recently secured fresh funding at a $965 billion valuation, surpassing OpenAI's $852 billion valuation reported in March.

Microsoft Restricts Employee Access to Anthropic's Claude Fable 5 Over Data Retention Concerns

According to reports, the decision stems from differences in how Claude Fable 5 handles user data compared to other Claude models currently available within Microsoft's ecosystem. A new artificial intelligence model from Anthropic has found itself at the center of an unusual policy divide inside Microsoft. Just days after Anthropic unveiled its flagship Claude Fable 5 model, Microsoft reportedly restricted employee access to the system, citing concerns related to data retention requirements. According to reports, the decision stems from differences in how Claude Fable 5 handles user data compared to other Claude models currently available within Microsoft's ecosystem. While the company continues to offer the model to external customers through its AI platforms, internal employees have reportedly been blocked from using it. Data Retention Rules Trigger Legal Scrutiny The key issue appears to be Anthropic's updated data retention framework. Most Claude models available to Microsoft employees operate under a Zero Data Retention (ZDR) arrangement, meaning user interactions are not stored for model training or operational purposes. Claude Fable 5 reportedly does not follow the same structure, leading Microsoft's legal and compliance teams to adopt a more cautious approach. Customer Access Remains Unaffected Expanding Microsoft-Anthropic Relationship The development comes despite an increasingly close relationship between Microsoft and Anthropic. In recent months, the companies have expanded collaboration around AI infrastructure and model availability. Anthropic's Claude family of models has been incorporated into multiple Microsoft offerings, including developer tools and AI-powered productivity solutions. Previous agreements reportedly included support for models such as Claude Sonnet 4.5, Opus 4.1, and Haiku 4.5 through Microsoft's AI ecosystem. The addition of Claude technology into enterprise platforms has positioned Anthropic as one of the key AI partners competing alongside other major model providers. AI Governance Becomes a Competitive Differentiator A Sign of the AI Industry's Next Phase The episode serves as a reminder that in today's AI landscape, technical performance alone may not be enough -- trust, transparency, and compliance are becoming critical competitive factors as well.

Microsoft Balks at Anthropic's Claude Fable 5 Data Retention Policy | PYMNTS.com

Anthropic's data retention policy for the newly released Claude Fable 5, which is a Mythos-class model, includes retaining prompts and outputs for 30 days on every platform on which the model is offered, and for up to two years if the prompts and outputs are flagged by Anthropic's trust and safety classifiers as violating the company's usage policy, according to the report. The company retains this data for trust and safety purposes, the report said. Anthropic announced Tuesday (June 9) that it launched Claude Fable 5 after developing safeguards to prevent it and other Mythos-class models from being misused. When the company announced the first Mythos-class model, Claude Mythos Preview, in April, it limited the release of the model to select partners so they could use its cybersecurity capabilities to strengthen their systems. Claude Fable 5 is safe for general use, with safeguards designed to prevent it from being misused for cybersecurity, biology and chemistry, or distillation, Anthropic said Tuesday. The company added that the model has capabilities that exceed those of any previous model Anthropic has made generally available, is state-of-the-art on most benchmarks of AI capability, and performs especially well in software engineering, knowledge work, vision and scientific research. In its Tuesday announcement of the release of the model, Anthropic said it introduced a new data retention policy for Claude Fable 5 and other models with similar of higher levels of capability. Anthropic said that the new 30-day data retention policy applies to both first- and third-party surfaces and that the company will ensure the data's deletion after 30 days in "almost all cases." The company added that it won't use this data to train new Claude models, or for any other purpose not related to safety, and that it logs all human access to the data. "The data will help us defend against complex and novel attacks (including new jailbreaks and attacks that operate across many requests) as well as help us identify and reduce false positives," Anthropic said in its announcement.

Why Anthropic's Fable 5 Marks the End of Free AI Services

Anthropic's latest release, Fable 5, represents a significant step forward in artificial intelligence, combining advanced reasoning capabilities with a strong focus on safety and ethical use. As detailed by Prompt Engineering, one standout feature is its ability to autonomously manage complex workflows, making it particularly valuable in fields like software engineering and genomics research. However, this performance comes with notable trade-offs, including increased computational demands and a pricing model that reflects its premium nature. These factors highlight the growing complexity and cost of deploying state-of-the-art AI systems. Explore how Fable 5 addresses challenges in cybersecurity, enhances scientific discovery and accelerates drug design through its adaptable and precise functionality. You'll also gain insight into the implications of its mandatory data retention policy and how its safety mechanisms balance ethical considerations with performance. This overview provides a clear breakdown of what organizations need to know to evaluate Fable 5's potential for their specific needs. What Makes Fable 5 Stand Out? Fable 5 represents the pinnacle of Anthropic's AI innovation, designed to handle intricate reasoning tasks with precision and efficiency. It incorporates advanced safety mechanisms to ensure responsible use, particularly in sensitive domains like cybersecurity. Methus 5, its less restricted counterpart, is tailored for specialized applications and is accessible only to a limited number of organizations. These models are poised to push the boundaries of AI's potential, offering solutions to some of the most demanding problems across industries. Key features that distinguish Fable 5 include its ability to process complex workflows autonomously, its robust ethical safeguards and its adaptability to a wide range of applications. These attributes make it a powerful tool for organizations seeking to use AI for innovation and operational efficiency. Performance Highlights Fable 5 has demonstrated exceptional performance, achieving a 46% score on the Frontier Code Benchmark, outperforming competitors like Opus 4.8 and GPT 5.5. This benchmark measures an AI model's ability to handle advanced coding tasks, showcasing Fable 5's superior reasoning and problem-solving capabilities. Its advanced reasoning capabilities enable it to tackle complex tasks with greater accuracy and reliability. However, this high level of performance comes at a cost: increased computational demands. Operating Fable 5 requires significant resources, reflecting the growing complexity of state-of-the-art AI systems. This trade-off between performance and resource consumption is a key consideration for organizations evaluating its adoption. Here are additional guides from our expansive article library that you may find useful on Claude Mythos. Key Applications Fable 5's versatility makes it a valuable tool across multiple domains, offering significant potential in areas such as: * Software Engineering: Automates code generation, debugging and optimization, streamlining development processes and reducing time-to-market for new applications. * Scientific Research: Assists in hypothesis generation and facilitates breakthroughs in molecular biology, genomics and other innovative scientific fields. * Cybersecurity: Enhances threat detection and response capabilities, providing organizations with robust tools to safeguard sensitive data and systems. * Drug Design: Accelerates the identification of potential compounds and optimizes development pipelines, contributing to faster and more cost-effective pharmaceutical innovation. These applications highlight Fable 5's ability to drive innovation and efficiency in industries seeking to use AI for complex problem-solving. Its adaptability ensures that it can address a wide range of challenges, from automating routine tasks to allowing new discoveries. Safety and Ethical Considerations Anthropic has placed a strong emphasis on safety in the design of Fable 5. The model includes advanced classifiers that redirect sensitive queries, particularly those related to cybersecurity, to less capable models like Opus 4.8. This approach minimizes the risk of misuse while maintaining robust functionality for legitimate applications. However, these safety measures can occasionally result in less responsive or degraded outputs for certain queries. This reflects the trade-offs involved in balancing security and performance. By prioritizing ethical considerations, Anthropic aims to ensure that Fable 5 is used responsibly, particularly in high-stakes environments where misuse could have significant consequences. Cost and Accessibility Fable 5 is one of the most expensive AI models on the market, priced at $10 per million input tokens and $50 per million output tokens. Access is provided through API and enterprise plans, with limited free usage available until June 2022. While the pricing reflects the model's advanced capabilities, it may pose a barrier for smaller organizations or individual users with limited budgets. The high cost underscores the premium nature of Fable 5, positioning it as a tool for organizations with the resources to invest in innovative AI solutions. For those able to afford it, the model offers immense potential to enhance productivity and innovation. Data Retention Policy To enhance safety and reduce false positives, Anthropic requires the retention of all traffic data for 30 days. This policy ensures that sensitive queries are appropriately handled and helps refine the model's safety mechanisms. Importantly, Anthropic has clarified that retained data will not be used for training new models or for purposes unrelated to safety, addressing potential privacy concerns. This approach reflects a commitment to transparency and accountability, balancing the need for robust safety measures with respect for user privacy. Organizations considering Fable 5 must weigh the benefits of its safety features against the implications of its data retention policy. Broader Implications The release of Fable 5 signals a significant shift in AI development, moving from task-specific assistance to autonomous systems capable of managing complex workflows. This evolution underscores the increasing costs associated with advanced AI models, marking the end of the "free lunch" era in AI services. For organizations, Fable 5 offers immense potential, but its pricing and data policies may require careful consideration before adoption. The model's capabilities highlight the growing role of AI in addressing global challenges, from advancing scientific research to enhancing cybersecurity. Challenges to Consider Despite its strengths, Fable 5 faces several challenges. Its high cost and mandatory data retention policies may deter smaller organizations or those with stringent privacy requirements. Additionally, the safety measures, while essential, can occasionally impact the model's responsiveness when handling sensitive queries. These factors highlight the complexities of deploying innovative AI systems in real-world scenarios. Organizations must carefully evaluate their needs and resources to determine whether Fable 5 aligns with their objectives and constraints. Looking Ahead Fable 5 marks the beginning of a new chapter in AI innovation, characterized by highly specialized models designed for narrow, well-defined tasks. As competition in the AI landscape intensifies, Fable 5 is expected to drive further advancements and inspire new applications across industries. While challenges remain, its launch underscores the fantastic potential of AI in addressing some of the world's most pressing problems. By allowing organizations to tackle complex challenges with greater efficiency and precision, Fable 5 paves the way for future breakthroughs in technology and beyond. Media Credit: Prompt Engineering Disclosure: Some of our articles include affiliate links. If you buy something through one of these links, Geeky Gadgets may earn an affiliate commission. Learn about our Disclosure Policy.

When Will Anthropic Release Mythos? - Apple (NASDAQ:AAPL), Amazon.com (NASDAQ:AMZN)

When Will Anthropic Release Mythos? Here's What Prediction Market Is Saying Anthropic is reportedly gearing up to make its advanced Claude Mythos AI model available to a larger audience. According to a Reuters report, Anthropic plans to soon offer Claude Mythos to all customers. This move follows the public release of its Claude Opus 4.8 model. Here's What Prediction Market Is Saying Data from Kalshi, a federally authorized betting platform, shows that over $390,000 has been bet on the contract "When will Anthropic release Mythos?" Bettors are highly confident that Mythos will be released "Before Aug 1, 2026," placing an 825 probability on the date. Some bettors are predicting an early release, placing a 71% probability for a launch "Before Jul 1, 2026." Some bettors are predicting an earlier launch, placing a 39% probability on a launch "Before Jun 15, 2026." Anthropic's Public Debut Companies Praise Mythos Moreover, companies like Cisco and Palo Alto Networks are expected to benefit significantly from Mythos AI, as highlighted by Jim Cramer. These companies are part of Anthropic's Project Glasswing initiative. Anthropic's shift towards recursive self-improvement in AI development further underscores the evolving nature of AI technology. Disclaimer: This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors. Photo courtesy: Shutterstock Market News and Data brought to you by Benzinga APIs To add Benzinga News as your preferred source on Google, click here.

Anthropic rolls out Claude Fable 5 and Claude Mythos 5 with improved vision, memory, and research abilities

Anthropic has introduced Claude Fable 5, a Mythos-class model designed for general use with safety safeguards, and Claude Mythos 5, a restricted-access version for cybersecurity defenders and scientific research partners. Both models share the same underlying system, with Mythos 5 operating under reduced safeguards in controlled environments. Claude Fable 5 and Claude Mythos 5 Mythos-class models sit above the Opus class in capability, following earlier models introduced through Project Glasswing. The first model in this tier, Claude Mythos Preview, was released in April. Model positioning and overview Claude Fable 5 is positioned as Anthropic's most capable generally available model. It shows strong performance across AI benchmarks, including software engineering, knowledge work, vision tasks, and scientific research. Performance improves further as task complexity increases, with larger gains on long and multi-step workflows. Claude Mythos 5 is built on the same underlying system as Fable 5 but is deployed with reduced safeguards for approved users. It is currently limited to cybersecurity defenders and selected infrastructure and research partners. Core capabilities and performance Software engineering Claude Fable 5 delivers strong improvements in software engineering tasks, including large-scale codebase migrations and system-wide changes. In early evaluations, it completed engineering work that previously required months in a significantly shorter time. It also shows improved efficiency on production-grade coding benchmarks, including Cognition's FrontierCode evaluation. In testing, Stripe reported that Fable 5 compressed months of engineering work into days, including a migration across a 50-million-line Ruby codebase. Knowledge work and reasoning The model performs strongly in structured reasoning tasks such as financial analysis, document interpretation, chart and table understanding, and multi-step problem solving. On Hebbia's Finance Benchmark, Fable 5 achieved leading performance in senior-level reasoning tasks. IMC evaluations reported strong results across trading analysis tasks, including factual lookup, conceptual reasoning, root-cause analysis, and expected-value analysis. Vision capabilities Claude Fable 5 improves visual understanding tasks, including extracting precise values from scientific figures, interpreting complex charts, and reconstructing application code from screenshots. It also requires less external scaffolding compared to earlier models. In testing scenarios, it successfully completed tasks such as playing Pokémon FireRed using a minimal vision-only setup, where earlier models required additional tooling. Memory and long-context processing The model supports extended context handling across multi-million-token inputs while maintaining coherence in long-running tasks. It improves outputs using persistent memory and self-generated notes. In experimental gameplay scenarios such as Slay the Spire, adding persistent file-based memory improved performance three times more than earlier models, with significantly higher success in reaching later stages. Scientific research and advanced applications (Mythos 5) Claude Mythos 5 is used in controlled scientific environments and demonstrates advanced capabilities in biological and computational research, including: * Protein and drug design workflows * Binding site selection and bioinformatics tool execution * End-to-end autonomous research task execution * Large-scale genomics analysis across millions of cells and 138 animal species * Generation of novel biological hypotheses In evaluations, Mythos 5 performed multi-step research workflows with minimal human input and matched or exceeded skilled human operators in specific tasks. Novel research outcomes Mythos 5 has contributed to several scientific findings, including: * Molecular biology hypotheses, with some advanced to experimental testing * A validated hypothesis involving a novel mechanism in an E. coli protein * Genomics research using single-cell datasets across 138 animal species * Machine learning models identifying equivalent cell types across distantly related organisms * Research outputs that in some cases outperform recently published models, including a model published in Science Some results are planned for publication, while others remain under active evaluation. Alignment and behavior assessment Internal assessments indicate that Mythos 5 shows low levels of misaligned behavior, including reduced deceptive outputs and limited cooperation with misuse scenarios. Its behavior is reported to be similar to earlier Opus-class models. Because Fable 5 is built on the same underlying system, its alignment characteristics are expected to be similar. Safety systems and safeguards Claude Fable 5 includes expanded safety systems designed to reduce misuse risks while maintaining usability. Safety classifier system The model uses AI-based classifiers to detect sensitive or potentially harmful requests. When triggered, responses may be routed to Claude Opus 4.8 instead of Fable 5. These systems cover cybersecurity, biology and chemistry, and model distillation-related risks. Most user sessions do not trigger fallback behavior. Cybersecurity safeguards Fable 5 includes protections against offensive cybersecurity use cases such as vulnerability exploitation and multi-step attack workflows, including reconnaissance, discovery, and lateral movement. Extensive red-teaming, internal evaluations, and external bug bounty testing report strong resistance to jailbreak attempts, including attempts using known public techniques. Biology and chemistry safeguards Stronger filtering is applied to sensitive biological and chemical research tasks due to increased model capability in these domains. These safeguards are broader than earlier systems and may occasionally flag benign requests. A tested example includes adeno-associated virus (AAV) design tasks, where models demonstrated strong predictive ability in genetic modification impact analysis, highlighting both scientific potential and dual-use risk. Distillation protection Systems are designed to detect attempts to extract or replicate model behavior for training competing models. Flagged requests may be redirected to safer responses. Data retention policy Anthropic applies a 30-day retention policy for business traffic involving Mythos-class models. * Data is not used to train new models * Access is strictly logged and monitored * Data is deleted after 30 days in most cases * Retention is used only for safety monitoring and abuse prevention Availability, pricing, and subscription rollout Claude Fable 5 is available broadly through API access and subscription plans with staged rollout based on capacity. During the initial rollout phase, it is included in Pro, Max, Team, and Enterprise plans at no additional cost. After this period, access will move to a usage-credit model, with possible extensions depending on capacity. Long-term plans aim to restore Fable 5 as a standard subscription feature once infrastructure stabilizes. Claude Mythos 5 remains restricted to trusted users, including cybersecurity partners and selected research organizations, with expansion planned through structured access programs for both cybersecurity and biomedical research. Both models are priced at:

Claude Fable 5: The Powerful AI Model That Anthropic Released with Guardrails in Place

Anthropic is releasing the version a day after arch rival OpenAI decided to file papers for an IPO, in what has become a race to supremacy between the two companies Anthropic today revealed that it has released its most powerful AI model to the general public and it comes with adequate guardrails. The Claude Fable 5 became the first publicly available version of what was till now a mysterious Mythos model for the world, barring those select few who got access under Project Glasswing. Per the Claude-maker, the new model excels at software engineering, knowledge work, and vision but has adequate guardrails in place to ensure that it does not breach certain safety limits predefined by Anthropic. For example, in areas like cybersecurity, biology, chemistry, and distillation, the model blocks responses and delivers Claude Opus 4.8 results. In recent times, Claude Mythos attracted attention for all sorts of reasons after Anthropic talked it up via early leaks and then direct responses. When a preview version was launched in April, the company itself limited access to a slew of partners over cybersecurity concerns while claiming that the frontier model was its smartest to date. This heightened interest from enterprises and security agencies but led to concerns amongst governments (including in India) who access directly or through the good offices of the Trump administration. A week ago, Anthropic expanded access to hundreds of organisations across 15 countries that included India and specifically to companies managing critical infrastructure.

Anthropic Launches Mythos-Class Models After Developing Safeguards | PYMNTS.com

Claude Fable 5 is safe for general use, while Claude Mythos 5 will initially be released for only a small group of cyber defenders and infrastructure providers through Project Glasswing, according to the release. The launch of these models follows the limited release in April of the first Mythos-class model, Claude Mythos Preview. Anthropic introduced Project Glasswing at that time, saying it would offer select partners early access to Claude Mythos Preview so they could use the model's cybersecurity capabilities to strengthen their systems. By May 22, Anthropic reported that Claude Mythos Preview had identified more than 10,000 cybersecurity vulnerabilities in "the most systemically important software in the world" so that they could be patched. Anthropic said in the Tuesday press release that since April, it has been developing safeguards for Mythos-class models. "Over the past few months we have been improving these safeguards, and they are now robust enough for a general release," Anthropic said in the Tuesday press release. The company said that it has prioritized safety, so users may encounter false positives from the safeguards. Claude Fable 5 has capabilities that exceed those of any previous model Anthropic has made generally available, is state-of-the-art on most benchmarks of AI capability, and performs especially well in software engineering, knowledge work, vision and scientific research, according to the release. Because Fable 5's capabilities could be misused, it has safeguards that will route queries on some topics to the less-capable Claude Opus 4.8. Those topics include cybersecurity, biology and chemistry, and distillation, per the release. Claude Mythos 5 is the same underlying model as Fable 5 but with cyber safeguards lifted, according to the release. Anthropic said Mythos 5 has "the strongest cybersecurity capabilities of any model in the world" and will initially be deployed through Project Glasswing in collaboration with the U.S. government. The company plans to expand access through a trusted access program. Mythos 5 was made available Tuesday to users who have access to Claude Mythos Preview. It is comparable to, or somewhat stronger than, that model but costs substantially less. "The capabilities of models like Fable 5 and Mythos 5 have the potential to do profound good for the world," Anthropic said in the release. "We've seen the beginnings of this in Project Glasswing, where the models have helped cyber defenders secure critically important software. We've also seen it in life sciences research, where the models are positing novel hypotheses and speeding up the development of new therapeutics."

Anthropic's Fable 5 draws mixed reactions from early users

Anthropic's new AI model, Claude Fable 5, is here. Early users see big improvements in handling complex tasks like software development and design. Experts praise its advanced capabilities. However, concerns are rising about its cost and how easily people can access it. The model is currently in subscription plans but will shift to usage-based pricing soon. Anthropic's newly released AI model, Claude Fable 5, is receiving a mixed response from early users, with developers and industry leaders highlighting both major improvements in capability and concerns around cost and access. Launched alongside Mythos 5, the model is designed to handle complex, long-running tasks such as software development, research and design. According to Anthropic, Fable 5 is a safeguarded version of Mythos, intended for broader public use. Early reactions on X (formerly Twitter) point to a noticeable jump in performance. Andrej Karpathy, an AI researcher and former OpenAI and Tesla executive, described the release as "a major-version-bump" and said the model performs strongly on benchmarks and in extended problem-solving sessions. Similarly, Garry Tan, CEO of Y Combinator, described the model as having "the biggest model energy" he has seen. Other users highlighted practical applications. Asaad Mahmood, founder of design agency The Small Square, said he used Fable 5 to design and launch a website, noting that the output showed structured design elements such as hierarchy and spacing typically associated with experienced human designers. He added that the gap between AI-generated and human design work is narrowing. Marc Andreessen, cofounder of Andreessen Horowitz, pointed to the model's potential in content generation, sharing an example prompt that asks the system to generate large volumes of editorial-style writing with high accuracy. Use cases highlighted by early testers include real-time software generation during customer calls, automated report and presentation creation, and pixel-accurate design replication. Deedy Das, an investor and engineer, cited examples ranging from generating consulting-style reports to building interactive 3D environments and UI systems. Ethan Mollick, a professor at the Wharton School who studies AI adoption, said the model can process large design documents and work for extended periods up to several hours before producing outputs. Similarly, Dan Shipper, CEO of Every, a media and software company, reported that internal testing showed the model completing large coding tasks autonomously, including clearing production bug backlogs and generating full applications in a single run. Boris Cherny, a software engineer, said the model demonstrates more structured debugging behaviour, including iterative testing and verification, without explicit prompting. The limitations Felix Rieseberg, who leads Claude Code and desktop AI initiatives at Anthropic, said the launch marks a broader transition in how AI systems are used from handling discrete tasks to managing ongoing "responsibilities." The company has initially rolled out Fable 5 within existing subscription tiers but with usage limits. Amol Avasare, head of growth at Anthropic, said the model will be included in subscription plans for a limited period approximately two weeks after which access will shift toward usage-based pricing via token consumption. Some developers note that the model's higher capability comes with increased computational cost. Shipper said Fable 5 is significantly more expensive than earlier models and consumes substantially more tokens per task, making it better suited for high-complexity workloads rather than routine use. Bindu Reddy, CEO of Abacus.AI, said internal evaluations suggest performance gains are most evident in a small percentage of highly complex coding tasks.

Anthropic opens Mythos-class AI to the public with data retention

* Read Anthropic's Mythos announcement here. Anthropic has launched two Mythos-class models on June 9, 2026: Claude Fable 5, the first such model it has made publicly available, and Claude Mythos 5, restricted to vetted cybersecurity partners. The two run on the same underlying model, and Anthropic attached a mandatory 30-day data-retention requirement to both, overriding existing zero-retention agreements. The two models, side by side: Both run on identical weights, capabilities, and pricing. What separates them: * Claude Fable 5: Public. Anyone can use it, with classifiers that intercept high-risk prompts and route them to the less capable Claude Opus 4.8 model. * Claude Mythos 5: Restricted. Goes only to Project Glasswing cyber defenders and critical-infrastructure partners, with cybersecurity safeguards lifted. Access depends on who the user is, not on a paid upgrade. Anthropic will retain all Mythos-class traffic for 30 days: Anthropic will hold all traffic on both models for 30 days, on its own and third-party surfaces, even for enterprises that previously had zero-retention deals. It said it will: * Use the data only to defend against novel attacks and reduce false positives, not to train models. * Log all human access to the data and delete it after 30 days in almost all cases. The policy could set an industry precedent where access to frontier models comes bundled with mandatory retention framed as a safety measure, a direct concern for Indian enterprises and developers on the Application Programming Interface (API). How the safeguards work: Fable 5 sits behind a layer of classifiers, separate Artificial Intelligence (AI) systems that screen prompts before they reach the model. * When a request touches cybersecurity, biology, chemistry, or model distillation, Fable 5 routes it to Opus 4.8 rather than answering. * Anthropic said more than 95% of sessions trigger no fallback, so most users get performance equal to Mythos 5. * In MediaNama's own testing on June 10, Fable 5 routed a routine request to audit an open-source web application to Opus 4.8, an example of the classifiers catching standard defensive security work. Screenshot: Fable 5 switched a routine request to audit an open-source web application to Opus 4.8, rather than answering directly. (MediaNama testing, June 10, 2026) Anthropic says it could not break its own safeguards: Anthropic said it ran over 1,000 hours of external testing and found no universal jailbreak, a method that lets a user bypass a model's safeguards entirely, though it acknowledged the United Kingdom's AI Security Institute made partial progress towards one in a brief window. The claim matters because Anthropic has struggled to contain the technology before: on the original Mythos model's launch day, a group exploited contractor credentials to reach it through a third-party vendor environment. India recently gained restricted access: Anthropic deploys Mythos 5 through Project Glasswing in collaboration with the United States government and plans a broader trusted-access program. India was shut out of the program through April, but Anthropic's mid-2026 expansion to over 15 countries brought in a single-digit number of Indian entities, spanning government and private sectors, according to sources cited by Livemint. Neither Anthropic nor the organisations have disclosed which Indian institutions hold the access. Fable 5 now offers the wider Indian public the capability in guard-railed form, while Mythos 5 stays limited to that small, unnamed cohort. Why it matters: Fable 5 makes the most capable model Anthropic has released publicly available, but on terms that tighten control rather than loosen it. MediaNama founder and editor Nikhil Pahwa has argued in Reasoned that a tool which "compresses attack timelines without compressing defence timelines increases systemic risk before it improves security," and that logic extends to a public release. Wider access spreads both the defensive and offensive potential of Mythos-class capability, and the retention mandate signals that Anthropic intends to monitor how the public uses it. Pricing: Fable 5 and Mythos 5 cost $10 per million input tokens and $50 per million output tokens, which Anthropic said is less than half the price of the earlier Mythos Preview. Anthropic will include Fable 5 in Pro, Max, Team, and seat-based Enterprise subscription plans until June 22 and then require usage credits, with a plan to restore it as a standard subscription feature when capacity allows.

Claude Mythos 5 and Fable 5 Officially Launch

Anthropic's latest AI models, Claude Fable 5 and Claude Mythos 5, mark a significant step forward in artificial intelligence development. As highlighted by Nate Herk, these models belong to the new Mythos class, which introduces distinct capabilities tailored to different user needs. Claude Fable 5, included temporarily in subscription plans like Pro Max and Enterprise, offers a versatile solution with robust safeguards for general use. On the other hand, Claude Mythos 5 is reserved for select partners under Project Glasswing, featuring fewer restrictions and advanced functionality for high-stakes applications such as cybersecurity and scientific research. Both models are priced at double the cost of the Opus series, reflecting their enhanced performance and specialized features. Explore how these models excel across various domains, from improving reasoning in knowledge work to advancing software engineering and data verification. Gain insight into the practical applications of Claude Fable 5, particularly for users using its temporary availability in subscription plans. Additionally, understand how Claude Mythos 5 is being deployed under strict ethical guidelines to address complex challenges while prioritizing safety. This guide provides a detailed breakdown of their capabilities, pricing and potential impact across industries. Key Features of the Mythos Class The Mythos class introduces two distinct models, each designed to meet specific user needs: * Claude Fable 5: A versatile, general-use model equipped with robust safeguards to ensure responsible usage. Temporarily included in Pro Max, Team and Enterprise subscription plans, it offers a wider audience the opportunity to explore its capabilities. * Claude Mythos 5: A innovative model with fewer restrictions, delivering unmatched performance. Access is limited to select partners under Project Glasswing to ensure its deployment aligns with safety and ethical standards. Positioned above the Opus series, these models set a new standard in AI technology, excelling in reasoning, analytical tasks and autonomous operations. Their design reflects Anthropic's commitment to delivering tools that balance innovation with responsibility. Pricing and Limited Availability The Mythos class models are priced at $10 per million input tokens and $50 per million output tokens, doubling the cost of the Opus series. For a limited time, Claude Fable 5 is included in Pro Max, Team and Enterprise subscription plans until June 22, 2026. After this date, access will transition to a token-based usage credit system. This temporary inclusion allows users to explore the model's capabilities and assess its value before adapting to the new pricing structure. Uncover more insights about Claude Mythos in previous articles we have written. Enhanced Performance and Applications Claude Fable 5 and Mythos 5 deliver substantial improvements over their predecessors, excelling in a variety of fields. Their advanced capabilities make them particularly effective in: * Software Engineering: Offering sophisticated coding tools and agentic programming support. * Knowledge Work: Enhancing reasoning and analytical abilities for tackling complex problems. * Vision Tasks: Supporting creative projects and improving data verification processes. * Scientific Research: Contributing to new studies and fostering innovation. * Cybersecurity: Providing robust tools to identify vulnerabilities and mitigate risks effectively. These advancements make the Mythos class indispensable for professionals across industries, allowing them to streamline workflows, solve intricate challenges and drive innovation with precision. Focus on Cybersecurity and Safety Anthropic has placed a strong emphasis on safety and cybersecurity in the development of the Mythos class. Claude Mythos 5, in particular, incorporates the most advanced cybersecurity features, but its access is restricted to trusted partners to prevent misuse. Both models are equipped with safeguards designed to mitigate risks, making sure their powerful capabilities are deployed responsibly. This focus on safety underscores Anthropic's commitment to ethical AI development, addressing the dual-use nature of such advanced systems. Future Expansion Plans Anthropic has outlined plans to expand the availability of Claude Fable 5 and Mythos 5 as operational capacity increases. Claude Fable 5 may return to subscription plans in the future, offering broader access to its advanced features. Additionally, Anthropic is collaborating with the U.S. government on a trust access program for Claude Mythos 5, making sure its deployment adheres to stringent safety and security standards. These initiatives reflect Anthropic's vision of making advanced AI tools accessible while maintaining a focus on ethical use. Real-World Use Cases The practical applications of Claude Fable 5 and Mythos 5 are extensive, making them ideal for tasks that demand high levels of reasoning and analytical performance. Key use cases include: * Software Development: Assisting developers with coding, debugging and optimizing software solutions. * Data Verification: Analyzing and validating textual and visual data with precision. * Scientific Research: Supporting researchers in designing experiments and interpreting complex results. * Cybersecurity: Identifying system vulnerabilities and implementing effective countermeasures. These models empower professionals to achieve greater efficiency and accuracy in their work while adhering to strict safety protocols to prevent misuse. Performance Insights The Mythos class models deliver exceptional performance, particularly in reasoning, analytical tasks and vision-based applications. Their enhanced capabilities make them well-suited for creative projects, data verification and complex problem-solving. By setting a new benchmark in AI technology, the Mythos class offers tools that can transform industries while maintaining a focus on ethical and responsible use. What You Should Consider If you are considering exploring Claude Fable 5, take advantage of its temporary inclusion in subscription plans before June 22, 2026. After this date, the shift to a token-based usage credit system will require users to adapt to a new pricing structure. By using the advanced features of the Mythos class models, you can enhance productivity, tackle complex challenges and drive innovation, all while adhering to the safety measures designed to ensure responsible use. Media Credit: Nate Herk | AI Automation Disclosure: Some of our articles include affiliate links. If you buy something through one of these links, Geeky Gadgets may earn an affiliate commission. Learn about our Disclosure Policy.

Anthropic Launches Claude Fable 5 with Advanced Safety Controls

Anthropic has unveiled Claude Fable 5, its most advanced public AI model yet. The launch combines stronger reasoning and coding capabilities with new safety measures designed to limit misuse in high-risk domains. Anthropic has introduced Claude Fable 5, the first publicly available model from its advanced Mythos family. The release marks a notable shift for the company, which had previously suggested that systems at this level of capability posed risks that made wider deployment difficult. The company now believes those concerns can be addressed through a combination of technical safeguards, monitoring systems, and controlled access policies.

Anthropic rolls out public version of Mythos without cybersecurity capability

Anthropic is rolling out a public version of its Mythos AI model, but with guardrails barring its use in risky areas such as cybersecurity, after a preview earlier this year sent shockwaves globally with its ability to find software flaws. The new Claude Fable 5 is the most powerful model Anthropic has ever made for wider use, the startup said on Tuesday, touting its performance in software engineering and analytics. Anthropic has so far limited its access to a group of about 200 organizations including the U.S. government under the Glasswing program, after announcing in April that Mythos had uncovered thousands of software vulnerabilities. Offering its capabilities more widely may allow the $965 billion company to extend the momentum that has powered its valuation above rival OpenAI just as the two startups at the center of the AI industry race to go public. The company said it had done extensive testing to ensure that users could not manipulate the new model to bypass its guidelines and perform restricted actions. "Let's say I'm a college student asking the model like help me find cyber vulnerabilities on X package or code. The model would refuse and Fable 5 will fall back to Opus 4.8 for a response," Dianne Penn, Anthropic's head of product management, research and labs, told Reuters. Fable 5 will be a more expensive model, but it accomplishes tasks with lower token usage, bringing the overall cost per task down, according to early customer feedback, Penn said. Anthropic also said users who had access to the preview version of Claude Mythos, the version without guardrails, would be able to upgrade to the new Claude Mythos 5. The company said it planned to expand access over time through a more "systematic trusted-access program." Pricing on both models is US$10 per million input tokens and $50 per million output tokens, the company said.

Anthropic, which claimed AI model was too risky for public to use, releases 'safe' version

After hyping potential safety threats posed by its latest AI model, Anthropic says it is releasing a version with tweaks that make it "safe" for the general public. A day after filing for its initial public offering, the company on Tuesday trumpeted a new version of the Mythos model with guardrails around cybersecurity and biological weapons. If the model, called Claude Fable 5, is asked about off-limits issues like exploiting a software bug or creating bioweapons, the response will be blocked and users will be directed to an older model known as Opus 4.8, Anthropic said in a press release. The company, led by Chief Executive Dario Amodei, has for years hyped the dangerous potential of its products while painting itself as a responsible steward of powerful AI technology. The San Francisco-based firm is on course for competing public listings this year with rivals OpenAI and Elon Musk's SpaceX, which merged with his AI outfit xAI in February. Anthropic is calling the new model "Mythos-class," a tier of Claude models which is more capable than its prior "Opus-class" models. "Releasing a model this capable comes with risks. Without safeguards, Fable 5's capabilities in areas like cybersecurity could be misused to cause serious damage," Anthropic said. "We've therefore launched the model with safeguards that mean queries on some topics will instead receive a response from our next-most-capable model, Claude Opus 4.8." The company went on to explain that "harmless" requests will sometimes get blocked, but said that would be rare. "We wanted to be able to provide this level of intelligence for general users in a safe manner," Anthropic's head of product management Dianne Penn told the Wall Street Journal. Anthropic spooked the tech industry and governments when it announced earlier this year that Mythos was powerful enough to cause widespread cybersecurity disruptions. It decided to limit the product's release to about 200 organizations. Anthropic has sounded a steady drumbeat of caution about AI, drawing skepticism from some critics. Just last week, the company called for an industry-wide pause to slow the pace of development while the industry gets a handle on potential societal risks. Commentators and rivals said Anthropic just wants to hobble the competition in the red-hot race to develop advanced artificial intelligence.

Anthropic unveils Claude Fable 5, making a shift toward autonomous AI agents

Anthropic has launched Claude Fable 5 and Claude Mythos 5, introducing a new class of AI models designed for long-horizon, autonomous tasks. The release highlights the industry's shift toward agentic AI, combining breakthroughs in software engineering and scientific research with stringent safeguards aimed at mitigating emerging cybersecurity and biosecurity risks. San Francisco-based AI company Anthropic has introduced Claude Fable and Claude Mythos 5, a new generation of "Mythos- class" models that signal broader industry shift from conversational AI to autonomous systems capable of executing complex, long-duration tasks. The announcement comes as competition among AI developers increasingly focuses on agentic capabilities rather than conventional chatbot performance. While Claude Fable 5 is being made broadly available through the company's API, Claude Mythos 5 will be offered under restricted access programmes for selected organizations and government collaborations, including Project Glasswing. According to Anthropic, the new models represent its most advanced AI systems to date. What sets them apart is their ability to sustain reasoning and execute multi-step workflows over extended periods, allowing them to tackle tasks that previously required significant human oversight. Beyond coding, the models have demonstrated notable gains in vision capabilities. Anthropic said the systems can interact with complex video games using only screenshots, without relying on additional game-state information or navigation tools. In scientific research, the company reported that the models have generated novel hypotheses in molecular biology and genomics, with some findings already receiving independent validation. The launch also underscored why Anthropic has adopted a differentiated deployment strategy. The company believes Mythos-class models possess capabilities that could potentially be misused in areas such as cybersecurity and biological research. To mitigate these risks Claude Fable 5 incorporated conservative safeguards that automatically redirect high-risks queries to the older Claude Opus 4.8 model. Anthropic acknowledged that this approach could result in occasional false positives but said the safeguard would be refined over time. As part of its safety framework, the company has also introduced a 30-day data retention policy for Mythos class traffic to study emerging attack patterns and strengthen its protective systems. Both models are priced at $10 per million input tokens and $50 per million output tokens, with access to Fable 5 being rolled out in phases across Pro, Team and Enterprise plans through June 2026. The launch highlights a broader transition in artificial intelligence from systems designed primarily to answer questions to AI agents capable of planning, executing and validating complex tasks with increasing levels of autonomy, while raising fresh questions about how such capabilities should be governed. Nominate for most innovative AI product 2026 Disclaimer Statement: This content is authored by a 3rd party. The views expressed here are that of the respective authors/ entities and do not represent the views of Economic Times (ET). ET does not guarantee, vouch for or endorse any of its contents nor is responsible for them in any manner whatsoever. Please take all steps necessary to ascertain that any information and content provided is correct, updated, and verified. ET hereby disclaims any and all warranties, express or implied, relating to the report and any content therein.

Anthropic opens most powerful AI model to public with safeguards

Anthropic, maker of the Claude artificial intelligence (AI) models, on Tuesday made the most powerful version of its technology available to the general public while restricting it from use in sensitive areas. most queries to it relating to sensitive issues like cybersecurity will be routed instead to the lower-tier model, Opus 4.8, which was made public in late May. Anthropic, maker of the Claude artificial intelligence (AI) models, on Tuesday made the most powerful version of its technology available to the general public while restricting it from use in sensitive areasDubbed Fable 5, the model is the first from the Mythos class -- Anthropic's most advanced lineup of AI technology, unveiled in April but restricted over cybersecurity concerns -- to be made widely available. Anthropic says Fable 5 is exceptionally good at writing and debugging software code, answering complex research questions and analyzing images. In parallel, Anthropic is offering an unrestricted version, Claude Mythos 5, to companies and organizations that already have access to this model family -- including cybersecurity partners enrolled in its Project Glasswing program. That select group was expanded in early June to around 200 organizations in more than 15 countries and is expected to grow further. Anthropic has restricted access to Mythos on cybersecurity grounds, given what the company calls its ability to quickly identify vulnerabilities in critical infrastructure, such as banking, power grids, and software. When Project Glasswing launched, some critics accused Anthropic of overhyping the threat so as to generate attention. But companies that have tested Mythos have since endorsed its capabilities, with the US government -- which had been in a legal dispute with Anthropic -- eventually testing the model over security concerns. The White House has since set up an arrangement to test the most powerful models from the leading AI companies before they are released. Anthropic says the restrictions on Fable 5 will work like this: most queries to it relating to sensitive issues like cybersecurity will be routed instead to the lower-tier model, Opus 4.8, which was made public in late May. Anthropic hired outside experts to spend more than 1,000 hours trying to find ways to bypass these restrictions -- a process known as red-teaming. The company also ran a bug bounty program, which pays people to find security flaws. No one found a way to completely unlock the model, Anthropic said. The security question, which Anthropic has made one of its main commercial arguments, had drawn the San Francisco startup into an unprecedented standoff with the Trump administration over its refusal to lift its restrictions on mass surveillance and autonomous lethal weapons. In the wake of that dispute, the Pentagon severed its contracts with the company, whose AI tools had been the only ones to hold defense security clearance. The launch of Fable 5 comes with a steep price tag -- $10 per million input tokens and $50 per million output tokens, double the cost of Opus 4.8. Tokens are measures of usage of an AI model. An intensive coding session by a programmer can burn through one million tokens in a matter of hours or less. Despite exponential revenue growth, Anthropic remains far from profitability and is paying a premium for computing power. It recently began leasing a data center from Elon Musk for $1.25 billion per month. These launches come amid intense financial excitement around AI. Anthropic and rival OpenAI recently announced they had filed IPO plans, just days ahead of a record-breaking market debut for SpaceX, Elon Musk's space giant that also owns xAI, his AI company.

Microsoft is restricting Claude Fable 5 for staff despite its AI ambitions, here is why

Microsoft has not added Claude Fable 5 to the internal version of GitHub Copilot that employees use for work. Anthropic recently launched Claude Fable 5, the first Mythos-class AI model designed for public use. The company claims that Fable 5 offers similar capabilities as Mythos AI while adding safety measures. While Microsoft quickly made the model available to its customers, the company is reportedly limiting its use internally. According to a report by The Verge, Microsoft has not added Claude Fable 5 to the internal version of GitHub Copilot that employees use for work. Other Claude models are available because they operate under Zero Data Retention (ZDR) rules, which do not store prompts or responses. The reported restriction is linked to new data retention rules introduced with Claude Fable 5. Unlike previous Claude models, Fable 5 requires Anthropic to temporarily store prompts and outputs in order to operate new safety classifiers. Anthropic reportedly keeps the prompts and outputs data for up to 30 days before deleting it. Also read: AI may bring cyberattacks and job losses if left unchecked, warns Anthropic CEO Dario Amodei The situation becomes more complicated when Anthropic's safety systems flag content as a potential violation of its usage policies. In such cases, prompts and outputs may be stored for as long as two years. This has reportedly raised concerns inside Microsoft about how employee data, customer information and other confidential content could be handled. Microsoft has said to inform employees that its legal teams are currently reviewing Anthropic's data retention requirements. As of now, it is unclear if Microsoft will eventually approve the model for employee use. Interestingly, as soon as the Claude Fable 5 was announced, Microsoft quickly rolled out Claude Fable 5 to its GitHub Copilot and Foundry customers. So, it seems like that the tech giant is taking a more careful approach when it comes to internal adoption. Also read: Anthropic launches Claude Fable 5, a safer Mythos AI model: Features, availability and how to use it Claude Fable 5 arrives just weeks after Anthropic claimed that the Mythos AI is so capable at cybersecurity tasks that it was too dangerous to release publicly. As per Anthropic, Fable 5 performs better than its previous public AI models. The model can handle complex software projects, understand visual information, analyse large amounts of data, and even stay focused during long tasks involving millions of tokens.

Claude Fable 5 is the beginning of the end of 'one model for every use case': Here's how

I remember when the big AI story was whether these models could pass a bar exam. Now one of them is designing drug candidates without a human in the room. We moved fast. Yet there is an underlying change in Anthropic's release of Fable 5 and Mythos 5 that strikes me as more significant than any particular metric: the identical technology is sold as two distinct products based on your identity. Fable 5 is available to all. Mythos 5 is available to individuals who have been approved by the US government. Identical weights, but separate policies. Even in naming these products, Anthropic made clear what kind of distinction they are making - Fable and Mythos are derived from the same Latin word roots. Also read: No PhD required: Mythos 5 ran an entire drug design workflow by itself This is unprecedented. Not in having enterprise tiers, which every SaaS startup provides. What's unprecedented here is the rationing of the capability itself on the grounds of identity rather than any other criteria. It doesn't matter whether we give a slightly slower or less expensive version to non-approved entities; what matters is the restriction of certain capabilities. An approved cybersecurity expert will be able to think through the offensive exploit chain using Anthropic's technology. All others will have access to Opus 4.8. Of course, the immediate argument against this is that it is the responsible choice. Anthropic is right in pointing out that having Mythos class cybersecurity capability in the wrong hands is indeed dangerous. But that doesn't change the fact that this is the creation of a two-tiered world with respect to artificial intelligence, where the government and institutions of trust have one thing, while the rest of us have another. This divide will only continue to grow. The next version of Mythos will be even better, while the number of institutions that have been granted permission to use the API will grow slowly. At the same time, classifiers will continue to make mistakes while researchers in India struggle with the fallback system when the chosen biotech startup in San Francisco has their hands on the Mythos 5 API key. Also read: Claude Fable 5 is less risky Mythos model: Safest AI right now? Anthropic's launch post is full of genuinely impressive demos. Fable 5 beat Pokémon FireRed using only raw screenshots - no maps, no helper tools, just vision. Earlier Claude models needed a full harness to attempt the same thing. It's the kind of detail that makes for great copy, and Anthropic knows it. But that's exactly the point. The demos are designed to dazzle. The Pokémon playthrough, the Factorio factory, the fluid simulation set to music - they're all real, and they're all remarkable. They're also a distraction from the quieter, more consequential thing happening underneath: a credentialed hierarchy being dressed up as a safety policy. The "one model for everyone" era wasn't perfect. But it was legible. What's replacing it isn't. That's the real story of this launch. Not the benchmarks. Not the Pokémon playthrough. The doors that just got installed and who holds the keys.

Claude Fable 5 is less risky Mythos model: Safest AI right now?

Over 1,000 red-team hours found no universal jailbreaks before launch, claims Anthropic Anthropic's much-anticipated Mythos class model finally dropped, called as Claude Fable 5. Before getting into what makes this latest AI model from Anthropic so unique, it's also an industry-first of sorts in terms of safety. Claude Fable 5's unique offering isn't just safety that prohibits the misuse of Mythos, but how it becomes selectively less powerful when power itself is its biggest risk. According to official disclosure, Claude Fable 5 is the first model in Anthropic's Claude 5 family. It's part of a new "Mythos-class" tier that sits above Claude Opus in capability right now. Claude Fable 5 and Claude Mythos 5 share the same underlying model, for all practical and general purposes. However, the key difference is that Fable 5 has additional safety measures built in which allowed it to be released publicly - whereas Mythos 5 is the more unhindered, unrestricted, superpowerful cybersecurity nuke model which is still private. Anthropic only showed to cybersecurity companies and big tech platforms in April 2026 under Mythos Preview, because they hadn't figured out how to predictively enforce guardrails that prevented Mythos from being misused at that time. Fast forward by just a couple of months and Anthropic is confident Claude Mythos 5 can't be misused through Claude Fable 5. How exactly is Anthropic achieving this level of safety? By simply routing dodgy and dangerous requests made to Claude Fable 5 to lesser capable models like Opus and Sonnet. No flat refusal, just incomplete information in the knowledge bank to answer confidently. I'd like to imagine this safeguard in Claude Fable 5 as a Class 10 student asking a PhD-level AI model on how to cheat at an upcoming exam. When the PhD-level model detects the nature of the request as dangerous, unethical, unsafe, it sends the request to a lower-level AI model that isn't as smart enough to respond. According to Anthropic, answering certain cyber, biology, and chemistry related prompts will trigger Claude Fable 5 to pass them on to less-capable Claude Opus 4.8 - which doesn't have the reasoning depth nor the information stored to successfully answer these queries which pose any danger. Of course, Anthropic isn't revealing all the safeguards it's baked into Claude Fable 5's guardrails. When people ask the model for dangerous questions related to cyber or biology, the step down model's response will be clearly indicated. However, that won't be the case for anyone trying to use Claude Fable 5 to train and build their own LLM, for example. Not only will Claude Fable 5 not handle these requests, users won't be told which fallback model is responding to them. Anthropic says this is for the better. Also read: Claude Mythos finds 10000 bugs: Is Indian industry ready? The real test will be whether Claude Fable 5 can avoid two common safety failures - false positives, where harmless queries get wrongly blocked, and false negatives, where cleverly disguised harmful requests slip through. Before its release, Anthropic mentioned that external red-teaming and a 1,000-hour bug bounty failed to find universal jailbreaks in the new Fable 5 and Mythos 5 model. However, Anthropic is setting a potential industry precedent by mandating a 30-day data retention policy on all traffic for this new class of Mythos models - including for enterprise customers with previous zero-retention agreements. Anthropic says this new policy shift is strictly in keeping for defending against novel attacks rather than any future model training.

Using Claude Fable 5? Anthropic says some topics are too dangerous to discuss, here is why

Anthropic is keeping its most powerful AI tools available only to trusted users while it assesses potential risks. Anthropic has launched Claude Fable 5, a more secure version of its Claude Mythos 5 AI model, alongside a new set of safeguards aimed at limiting misuse. The company says the latest model is far more capable than its earlier systems, making stronger safety measures necessary. As part of the rollout, Anthropic has placed strict restrictions on discussions related to cybersecurity, biology, and chemistry, areas that could be linked to harmful activities. Whereas Claude Fable 5 could previously provide a direct response to these sorts of sensitive questions, it is now being replaced by a less advanced model that was built earlier. This decision follows increasing awareness in the AI industry of both technology and safety. What topics can Claude Fable 5 not discuss? In a blog post, Anthropic said the most significant changes in Claude Fable 5 relate to how it handles sensitive subjects. The company explained that the chatbot may decline to answer certain queries involving cybersecurity, biology, or chemistry. In some cases, those requests may instead be handled by Claude Opus 4.8. Also read: Salesforce lays off more employees across teams, AI to takeover: Report Why Anthropic is restricting certain topics Anthropic's decision is driven by the growing capabilities of its latest AI systems. Claude Fable 5 is built on the same foundation as Claude Mythos 5, the company's most advanced AI model, which is currently available only to a limited group of trusted users. Mythos 5 has shown notable improvements in cybersecurity-related tasks. Researchers recently used the model to analyse multiple lines of code and identify software vulnerabilities in macOS. These capabilities demonstrate the potential benefits of advanced AI systems, but they also raise concerns about how such tools could be misused. Anthropic says it is taking a cautious approach because highly capable AI models could enable users to perform activities that would otherwise require specialised expertise. Therefore, the company has decided to add more safeguard measures on sensitive topics as they continue to evaluate the impact of the technology. Also read: Meta quietly pulled facial recognition feature after it was discovered in smart glasses app, criticised Trusted access remains limited Anthropic acknowledges that information useful to researchers and cybersecurity professionals could also be dangerous if it falls into the wrong hands. For that reason, access to Mythos 5 remains restricted through Project Glasswing, a programme designed for approved cybersecurity experts. The company plans to expand trusted access in the future, including to certain life sciences organisations. For now, however, Anthropic says it is prioritising safety, reflecting the wider debate across the AI industry about how increasingly powerful AI systems should be deployed and shared with the public.

Anthropic launches Claude Fable 5, a safer Mythos AI model: Features, availability and how to use it

The company says Fable 5 offers similar capabilities as Mythos AI while adding safety measures. Anthropic has launched Claude Fable 5, the first Mythos-class AI model designed for general use. The company says Fable 5 offers similar capabilities as Mythos AI while adding safety measures to reduce the risk of misuse. Mythos AI gained attention after several organisations found that it could independently discover critical software vulnerabilities faster than humans. While this capability can help improve cybersecurity, it can also be misused by hackers. Here is everything we know about Anthropic Claude Fable 5 AI model. Anthropic Claude Fable 5: Key features According to Anthropic, Claude Fable 5 offers many of the same strengths as Mythos while operating under stricter safety controls. The company says requests related to cybersecurity, chemistry, biology and some AI model training activities are monitored by dedicated safety systems. Also read: Starlink India launch reportedly on hold over security concerns: Here is what happened Anthropic explains that when these systems detect a potentially sensitive request, the query is automatically routed to a less capable model, Claude Opus 4.8. Anthropic says this allows users to receive answers while reducing the chances of misuse. In a blog post, Anthropic said, "Releasing a model this capable comes with risks. Without safeguards, Fable 5's capabilities in areas like cybersecurity could be misused to cause serious damage." The company also claims that Fable 5 performs better than its previous public AI models. It is said to handle complex software projects, understand visual information, analyse large amounts of data, and stay focused during long tasks involving millions of tokens. Alongside Fable 5, Anthropic has also introduced Claude Mythos 5, an updated version of its restricted AI model. Mythos 5 will remain available only to a limited group of organisations. Also read: Sam Altman says letting AI automate everything will be dangerous and unfulfilling, here is why Anthropic Claude Fable 5: Availability Claude Fable 5 is available globally starting today. The model is available through the Claude API and consumption-based Enterprise plans. Anthropic is also offering temporary access to users on Pro, Max, Team, and seat-based Enterprise plans at no additional cost from now until June 22. Starting June 23, access to Fable 5 on these subscription plans will require usage credits. However, Anthropic says it may extend the free access period if enough capacity is available. The company also said it plans to bring Fable 5 back as a standard feature on subscription plans in the future.

Anthropic may release Claude Mythos for public this week: Here is why it is a big deal

Anthropic has not confirmed the launch, but the release could bring some of its latest AI features to more users. Anthropic is reportedly gearing up to publicly launch Claude Mythos. The news comes right after Apple's WWDC26 event and OpenAI filing for its IPO. With the public version of Mythos, the AI giant is expected to bring some of the most advanced AI capabilities from Anthropic to a wider audience. However, the report also claims that the public version of the software may differ from the one which is currently being tested by select partners. With Mythos, Anthropic is said to offer stronger reasoning abilities and improved performance on tasks that require sustained attention over longer periods. According to a report by Sources.news, the public version of Anthropic's new model, Mythos, could be released on June 10, 2026. Moreover, it's also reported that the software may not be known as Mythos, a name that has been associated with it during the testing phase. The new tool from Anthropic will include substantial guardrails and intentionally limit some cybersecurity capabilities available in a separate partner-focused version. Also read: Sam Altman says letting AI automate everything will be dangerous and unfulfilling, here is why The upcoming model is said to have made significant progress in handling long and complex tasks that require multiple steps. Anthropic is expected to highlight these improvements as one of the key features with the launch of the public version. The report comes shortly after Anthropic announced an expansion of Project Glasswing, a programme designed to help organisations identify vulnerabilities and prepare responses to cyber threats using the company's AI technology. Earlier this month, the company said the initiative would expand to 150 organisations across more than 15 countries. Anthropic first provided access to the Mythos Preview model in April to around 50 partners, including agencies within the U.S. government. The newly added participants include organisations from the power, water, healthcare, telecommunications, and hardware sectors. Also read: Hackers exploit Microsoft open-source software to steal AI developers passwords According to the AI firm, these sectors were chosen because a successful attack on their systems could have widespread consequences. The company said the common factor among participating organizations is that attacks on their codebases could affect millions of people. Anthropic has not publicly confirmed details of the reported launch. A company spokesperson declined to comment on the claims. If the report proves accurate, the release could offer a clearer look at how Anthropic plans to bring advanced AI capabilities to the public while keeping certain high-risk functions restricted.