Anthropic export restrictions push India toward Chinese AI models as US access remains uncertain

Anthropic's export restrictions are fueling Asia's AI boom

The US move has allowed Asian firms to establish themselves in a market dominated by a handful of US companies. Anthropic's export restrictions have opened a window of opportunity for Asian AI startups, as they are rapidly rolling out cybersecurity models for clients who no longer have access to Anthropic's most advanced model. Over the past few weeks, startups in Japan and China have unveiled new AI models positioned as alternatives to Anthropic's latest Mythos and Fable 5 systems, seeking to capitalise on growing enterprise demand for advanced cybersecurity tools. Among the first movers is Tokyo-based Sakana AI, which introduced Fugu, an orchestration-based cybersecurity model that the company says delivers performance comparable to Anthropic's Fable 5 on several industry benchmarks. In China, 360 Security launched Tulongfeng, an AI-powered vulnerability discovery model designed to compete with Mythos. Chinese AI startup Z.ai has also showcased cybersecurity improvements in its latest GLM models, signalling the region's growing ambitions in frontier AI. The launches come as US export restrictions on Anthropic's most capable cybersecurity models continue over national security concerns. The restrictions were intended to prevent sensitive. AI capabilities from reaching certain overseas markets. But companies still require AI systems capable of vulnerability detection, malware analysis, and security research. This has driven demand for alternatives that can deliver similar capabilities without the uncertainty surrounding exports. The developments also reflect the rapid maturing of Asia's AI ecosystem. Rather than competing only on cost, startups are increasingly targeting specialised domains such as cybersecurity, and building models designed for enterprise use cases. Some companies are taking alternative approaches, including orchestrating multiple AI models instead of training a single frontier-scale system, allowing them to achieve competitive performance with lower development costs. As governments increasingly treat advanced AI models as strategic technologies, export controls are beginning to reshape where innovation happens. Similar to the semiconductor industry, restrictions on access are encouraging countries and companies to accelerate investments in domestic AI capabilities. While Anthropic has reportedly restored limited access to some trusted organisations, broader restrictions remain in place, allowing Asian firms to establish themselves in a market dominated by a handful of US companies.

US' Prime AI Off Table, India Opts for Fine China

The US government's tighter controls on frontier AI models from Anthropic and OpenAI are pushing Indian companies towards Asian alternatives as they seek project continuity and the performance gap narrows. Chinese open-source models, which were already popular among startups due to lower costs, may now make their way to large enterprises looking to avoid disruptions to top-dollar artificial intelligence projects. "The choice is pragmatic, not political," said Nipun Kalra, managing director and senior partner at BCG. The trend is "driven by cost, strong performance and licences that let enterprises self-host," said Kalra. Last week, China's Ziphu AI launched the GLM-5.2 model, which sits quite close to Anthropic's Opus 4.8 on performance benchmarks. The release has sparked a 'DeepSeek-like' moment because GLM is priced at a fifth the cost of Opus. Another Chinese firm, 360 Security Technology, unveiled two AI security tools under the banner Yitian Tulong -- named after Chinese epic novel Tulongfeng -- that automatically discover software vulnerabilities, and Yitianzhen to automate cyber defence and incident response. Japan's Sakana AI launched Fugu, a multi-agent orchestrator system that works as a single model. Reports claim these Asian labs are head-to-head with Anthropic and OpenAI. A new JPMorgan analysis showed that Chinese models now cost 10-50 times less per token than frontier models. These models from companies including DeepSeek, Alibaba, Xiaomi, MiniMax and Moonshot now dominate what it describes as the industry's 'intelligence-per-dollar' frontier, where performance is balanced against operating costs, the report said. "We're seeing real momentum - but the shift is towards open-weight models, and by independent benchmarks most of the well performing ones today happen to be Chinese," BCG's Kalra said. He, however, said western closed models still lead in frontier workloads. "Most enterprises are going multi-model and tiered, not switching wholesale." A recent joint research by the Massachusetts Institute of Technology (MIT) and Hugging Face revealed that Chinese models accounted for 17.1% of global downloads in 2025, surpassing the US share of 15.8%. "On capability, Chinese open-weight models are becoming extremely impressive," said Nikhil Narendran, partner at law firm Trilegal and a tech policy thought leader. "They are far cheaper, can often be downloaded and self-hosted, and the gap with frontier systems is narrowing fast." He explained that open-weights can run inside an enterprise environment and therefore have a privacy advantage. Trust Factor But whether enterprises can trust Chinese open-source models remains an open question. "If India merely responds to American dependency by rushing into Chinese dependency, we have not achieved sovereignty. We have merely changed our landlord," Narendran said. This month, the US government restricted access of Anthropic's Fable 5 and Mythos 5 models to foreign nationals. Last week, it asked OpenAI to limit the preview rollout of its GPT-5.6 family - Sol, Terra and Luna - to a small group of government-vetted partners. "This is probably the first time that ethnicities have become part of geopolitical strategy in a way that determines who can access critical technologies, and that could have far-reaching implications for R&D, hiring, funding and global collaboration," said Abishur Prakash, an author and geopolitical strategist at Canada-based advisory firm The Geopolitical Business Inc. "To me, the bigger question is how this trickles into more division within the global economy and society." Experts said such restrictions underscore the need for India to build a sovereign AI strategy to ensure access to latest frontier models and strengthen domestic capabilities. "The real lesson of recent weeks is concentration risk: dependence on any single provider can change by policy overnight," BCG's Kalra said. "For India, the priority is optionality and sovereign capability." (This story has not been edited by economictimes.com and is auto-generated from a syndicated feed we subscribe to.)

India's AI ambitions hit as US curbs access to key Anthropic models

India's reliance on foreign-controlled advanced AI models, like Anthropic's, is under scrutiny after a US government directive suspended access. Experts warn this highlights a significant vulnerability, potentially hindering India's ambition to become an AI power. The incident underscores the need for stronger domestic capabilities and strategic infrastructure to avoid being merely an adopter of technologies controlled elsewhere, emphasizing the geopolitical implications of frontier AI. Anthropic's suspension of access to its advanced Mythos and Fable models after a US government directive has sparked a debate over whether India is becoming overly reliant on advanced artificial intelligence technologies that are controlled by foreign entities. Policy experts and researchers caution that India can no longer assume it will have continued access to essential AI capabilities. The US government directive came shortly after India had secured access to the company's highly guarded AI systems. Experts say the development is not merely about one AI model being withdrawn. Instead, it has exposed a deeper vulnerability in India's AI ambitions. "It is arguably the most visceral example of the risks of depending on frontier AI controlled outside India, because access vanished overnight by government order rather than commercial choice," said Subimal Bhattacharjee, a technology policy analyst. "The gap between what's economically rational today and what's strategically safe long term has rarely been so visible." According to Bhattacharjee, dependence on foreign AI infrastructure is not new. India has long relied on foreign chips, cloud infrastructure, and digital platforms. However, the Anthropic episode made an abstract risk feel immediate by showing how access to frontier AI capabilities can ultimately depend on decisions taken outside the country. The incident has also raised questions over whether India risks becoming an AI adopter rather than an AI power. "A country becomes an AI power when it possesses meaningful influence over the foundational layers of the AI stack," said Kazim Rizvi, policy analyst and founder of public policy think tank The Dialogue. Those layers include advanced models, compute infrastructure, semiconductor access, data ecosystems, and the ability to shape how AI systems are deployed, he said. Rizvi said, "The issue is not merely access to one Anthropic model, but the fact that a capability relevant to cyber defence, vulnerability discovery, and critical infrastructure security could be withdrawn through decisions taken outside India." He added that nations relying entirely on external providers remain vulnerable to export controls, geopolitical tensions, and shifting regulatory priorities. "Frontier AI should now be treated as strategic infrastructure, not merely as a commercial software service," he said. Sarang Nerkar, former researcher and founder of Innosapien Technologies, said the episode should serve as a wake-up call for policymakers and enterprises alike. "It is not access to data that is being restricted. It is the model itself. The algorithmic layer is being controlled from abroad, and that is something we should take very seriously," Nerkar said. He cautioned that organisations building critical systems around frontier AI models need to carefully assess their dependence on technologies they do not control. "Imagine if access were suddenly restricted to a model that an organisation had relied on for two years. Entire workflows and dependencies could be affected," he said. The episode has also raised concerns about the concentration of power in the global AI ecosystem. "The incident is less about Mythos or Anthropic and more about the structural reality of today's AI ecosystem," said Mishi Choudhary, founder of Software Freedom Law Center India. "A handful of companies and governments control the most advanced models, the compute infrastructure, and often the terms of access." Choudhary said the answer is not technological isolation but greater resilience through open-source AI, public-interest research, stronger domestic capabilities, and procurement policies that reduce dependence on any single provider. As access to frontier AI increasingly becomes a geopolitical issue, experts argue India will need significantly greater investments in research, compute infrastructure and domestic AI capabilities if it wants a meaningful role in shaping the next phase of the global AI race.

After Mythos suspension, India pushes the US for stable AI access

The United States has verbally assured India that access to advanced AI technologies, once granted, will not be withdrawn without notice, according to a Hindu report. India reportedly requested this commitment after the Trump administration unexpectedly suspended Anthropic's two most powerful models for all foreign nationals earlier this month. What India was actually asking: India requested clarification from the US government regarding its regulatory framework following Washington's suspension of Anthropic's Fable 5 and Mythos 5 models under an export control order. MeitY Secretary S. Krishnan, speaking at the second Pax Silica Summit in Washington on Thursday, stated, "We can't have abrupt cutoffs" if India is to integrate these tools into its digital infrastructure and public services. Such a disruption could halt long-term AI initiatives, disrupt digital platforms that depend on continuous AI functionality, and undermine developmental goals that require consistent access to technology. What the US said -- and what it didn't put in writing: Krishnan stated there was "an understanding," not a formal guarantee, that "access to technology, once it is provided, will not be cut off." He described this as an assurance rather than a treaty or binding framework. Krishnan added that India received clarity on the United States' perspective and its intention to maintain access to technology for "trusted partners" in the future. What actually happened to Anthropic's models: On June 12, the Commerce Department's Bureau of Industry and Security directed Anthropic to suspend access to its Fable 5 and Mythos 5 models for all foreign nationals, both within and outside the United States, due to national security concerns. As the company could not reliably verify user nationality, it disabled both models entirely. Anthropic received the order at 5:21 p.m. ET and stated that the government did not specify the national security concern. The directive applied to individuals outside the US as well as foreign nationals within the US, including Anthropic's non-US-citizen employees. Anthropic maintained it had no option but to disable the models for all users. The company clarified that access to its less powerful Claude models remained unaffected. The stated trigger was a reported jailbreak. Officials informed Anthropic that the decision was based on a report of a technique that could bypass Fable 5's safeguards, which are intended to restrict access to the advanced cybersecurity features of Mythos, the model underlying Fable 5. However, the trigger is disputed. Anthropic described the incident as a narrow misunderstanding and noted that similar behaviour is common in other deployed models. At least one security researcher who reviewed the work rejected the term "jailbreak," instead describing it as defensive research. India's wider concern: The assurance-seeking extended beyond Anthropic. Krishnan addressed the broader issue of supply chain resilience, stating, "If you become over-dependent -- and I think that's what geopolitics and also things like the COVID pandemic have taught us -- you don't become over-reliant on one source of supply. Therefore, you need a multiplicity of at least three or four reliable and trusted sources of supply for a variety of technologies." Both sides agreed on the importance of strengthening the global technology ecosystem by diversifying supply chains and minimising reliance on any single region or supplier. On the question of regulation, Krishnan was explicit: "Right now, it is still time for innovation. It's not yet time to actually look at regulation in the sector -- that is a position that India has held." The broader friction this exposed: The Mythos suspension was not an isolated incident. This announcement represented Anthropic's latest conflict with the US government. After negotiations between the two organisations failed, the Department of Defense designated Anthropic as a supply chain risk, a label typically reserved for foreign adversaries. As a result, defense contractors were required to certify that they would not use Claude models in military projects. Anthropic filed a lawsuit against the Trump administration to challenge the blacklisting, and litigation is ongoing. The company also challenged the US government directive that halted access to its advanced AI models, saying that the "pace of frontier AI advancement is blistering, and competitive ground lost during a suspension cannot be regained after the fact." Critics highlighted a contradiction in the administration's stance: while emphasising the need for the US to lead China in AI, it simultaneously banned Britain and other non-Americans from accessing its top models. Analysts also warned that this directive could prompt many Chinese-born AI researchers in US labs to return to China and might lead investors to question the reliability of American AI companies due to the administration's unpredictable AI policy.

India Seeks Long-Term Access to Anthropic's Fable 5 as US-India AI Talks Begin

India and the United States have started important discussions on the Anthropic Fable rollout. Both countries want advanced AI technology to reach trusted partners without creating security risks. Officials also want strong frameworks in place before these powerful AI models become widely available. Jacob Helberg, US Under Secretary for Economic Affairs, confirmed the ongoing discussions. He said both countries understand each other's concerns and are working toward a common goal. The talks focus on national security, critical infrastructure, and safe access to frontier AI systems.

Living on borrowed AI won't end well for India, warn experts

India's reliance on foreign-controlled advanced AI models, like Anthropic's, is under scrutiny after a US government directive suspended access. Experts warn this highlights a significant vulnerability, potentially hindering India's ambition to become an AI power. The incident underscores the need for stronger domestic capabilities and strategic infrastructure to avoid being merely an adopter of technologies controlled elsewhere, emphasizing the geopolitical implications of frontier AI. Anthropic's suspension of access to its advanced Mythos and Fable models after a US government directive has sparked a debate over whether India is becoming overly reliant on advanced artificial intelligence technologies that are controlled by foreign entities. Policy experts and researchers caution that India can no longer assume it will have continued access to essential AI capabilities. The US government directive came shortly after India had secured access to the company's highly guarded AI systems. Experts say the development is not merely about one AI model being withdrawn. Instead, it has exposed a deeper vulnerability in India's AI ambitions. "It is arguably the most visceral example of the risks of depending on frontier AI controlled outside India, because access vanished overnight by government order rather than commercial choice," said Subimal Bhattacharjee, a technology policy analyst. "The gap between what's economically rational today and what's strategically safe long term has rarely been so visible." According to Bhattacharjee, dependence on foreign AI infrastructure is not new. India has long relied on foreign chips, cloud infrastructure, and digital platforms. However, the Anthropic episode made an abstract risk feel immediate by showing how access to frontier AI capabilities can ultimately depend on decisions taken outside the country. The incident has also raised questions over whether India risks becoming an AI adopter rather than an AI power. "A country becomes an AI power when it possesses meaningful influence over the foundational layers of the AI stack," said Kazim Rizvi, policy analyst and founder of public policy think tank The Dialogue. Those layers include advanced models, compute infrastructure, semiconductor access, data ecosystems, and the ability to shape how AI systems are deployed, he said. Rizvi said, "The issue is not merely access to one Anthropic model, but the fact that a capability relevant to cyber defence, vulnerability discovery, and critical infrastructure security could be withdrawn through decisions taken outside India." He added that nations relying entirely on external providers remain vulnerable to export controls, geopolitical tensions, and shifting regulatory priorities. "Frontier AI should now be treated as strategic infrastructure, not merely as a commercial software service," he said. Sarang Nerkar, former researcher and founder of Innosapien Technologies, said the episode should serve as a wake-up call for policymakers and enterprises alike. "It is not access to data that is being restricted. It is the model itself. The algorithmic layer is being controlled from abroad, and that is something we should take very seriously," Nerkar said. He cautioned that organisations building critical systems around frontier AI models need to carefully assess their dependence on technologies they do not control. "Imagine if access were suddenly restricted to a model that an organisation had relied on for two years. Entire workflows and dependencies could be affected," he said. The episode has also raised concerns about the concentration of power in the global AI ecosystem. "The incident is less about Mythos or Anthropic and more about the structural reality of today's AI ecosystem," said Mishi Choudhary, founder of Software Freedom Law Center India. "A handful of companies and governments control the most advanced models, the compute infrastructure, and often the terms of access." Choudhary said the answer is not technological isolation but greater resilience through open-source AI, public-interest research, stronger domestic capabilities, and procurement policies that reduce dependence on any single provider. As access to frontier AI increasingly becomes a geopolitical issue, experts argue India will need significantly greater investments in research, compute infrastructure and domestic AI capabilities if it wants a meaningful role in shaping the next phase of the global AI race.

US says in talks with India on Anthropic Fable model rollout

India and the US are holding crucial talks on releasing advanced AI models, including those from Anthropic. Both nations are prioritizing a secure, gradual approach to ensure critical infrastructure like power grids remain safe. India has voiced concerns about potential abrupt cutoffs in AI access, seeking assurances for reliable technology supply from trusted partners. This dialogue underscores a deepening US-India technology partnership focused on managing AI risks while fostering innovation. The United States and India are engaged in high-level, "sensitive" discussions regarding the release of advanced artificial intelligence models, including those developed by Anthropic. US Under Secretary of Economic Affairs Jacob Helberg confirmed the ongoing dialogue, emphasising that both nations are prioritising a secure and collaborative approach to frontier AI technology. In an exclusive conversation regarding the rollout of models like Anthropic's Fable to ANI, Helberg stressed that the US is committed to a "gradual, measured approach." This strategy is designed to ensure safety not only for the United States but also for India and other "trusted partners" who rely on these technologies for critical infrastructure, such as power grids. Helberg highlighted the complex nature of these negotiations, noting that while the discussions involve "sensitive national security" matters that cannot be fully detailed for public consumption, the two countries are in strong alignment. "We continue to have ongoing conversations about this topic with our Indian friends. These are very sensitive national security discussions that are not quite right for public consumption. But I think both sides really understand each other's perspectives, our intention is very much to continue a gradual measured approach to how we release Anthropic's models in a way that is safe, both for ourselves but also for our Indian counterparts as well as all our trusted partners for our critical infrastructure, for our power grid and so we are going to continue making sure that we do the hardwork in having those conversations. ," Helberg added, underscoring the deep cooperation between Washington and New Delhi. The ongoing dialogue reflects the growing importance of the US-India technology partnership. As both nations explore the integration of advanced AI across sensitive sectors, they are working to establish frameworks that manage the risks associated with frontier models while fostering technological advancement. By maintaining these high-level conversations, both sides aim to safeguard their critical infrastructure from the potential disruptions that can accompany the rapid deployment of powerful AI systems. The talks come at a time when India is rapidly integrating AI into its digital ecosystem and governance systems, while prioritising stability and uninterrupted technological access. India has raised concerns over the risk of abrupt disruptions in access to frontier AI systems and their potential impact on long-term projects. Speaking on the sidelines of the 2nd Pax Silica Summit in Washington, S Krishnan, Secretary of the Ministry of Electronics and Information Technology (MeitY), said that India had expressed concerns over the reliability of access to advanced AI models such as Anthropic's Claude. "We sought an understanding of how exactly the US is looking at this particular aspect and what their concerns are and how, in the future, this could be a reliable source of technology, because if it is something which is to be used and made available, we can't have abrupt cutoffs. We were given an understanding of how the US looks at this particular issue and how, going forward, they will ensure that for trusted partners, access will not be an issue," Krishnan said. The discussions are taking place as India explores broader use of advanced AI systems across multiple sectors, while simultaneously seeking safeguards against sudden disruptions in access. Officials noted that such interruptions in foundational AI technologies could arise due to geopolitical developments, export control regulations, or commercial decisions by providers, potentially affecting long-term AI-driven initiatives and digital infrastructure. To address these concerns, India sought greater clarity on the US regulatory framework and its long-term approach to AI technology access. In June, the US Commerce Department issued an export control directive requiring Anthropic to restrict foreign nationals from using its newly launched AI models, Claude Fable 5 and Mythos 5. Krishnan said India was given an understanding of how the United States views the issue and how it intends to ensure that access to technology does not become a challenge for trusted partners in the future. He added that India remains committed to strengthening international cooperation while promoting resilient and diversified technology supply chains to support the long-term growth of the global AI ecosystem.