Alice Secures $140M to Defend AI Models Against Infinite Attack Vectors

Schwartz: 'There are infinite ways to break an AI'

Alice, the Israeli firm that red-teams models for Anthropic, Google and Cohere, has raised $140mn led by Apax Digital. Its own announcement gives no valuation. According to Bloomberg it is worth close to $1bn, while Israeli outlets put it at $700mn to $800mn. One company spent eight years cataloguing the worst material on the internet. It has now raised $140mn to point that archive at AI models. Alice, formerly ActiveFence, announced the round on Tuesday. Apax Digital Funds led it, and will take a board seat. Total funding now stands at $280mn. The announcement names MoreTech and Phoenix Financial as the new participants. Existing backers including Resolute Ventures, Grove Ventures, CRV, Highland Europe, Norwest, NFX and Claltech also joined. Bloomberg and the Israeli press add two names the release leaves out. Samsung Electronics took part, and so did the listed cybersecurity company SentinelOne. That last one now holds a stake in a firm selling into the same buyers. Nobody agrees what it is worth The company's own announcement gives no valuation at all. Chief executive Noam Schwartz told Marissa Newman at Bloomberg that the round values Alice close to $1bn. Meir Orbach put it lower in Calcalist, at $700mn to $800mn. Globes reported $800mn. The gap between the low figure and the high one runs to roughly $300mn. That is more than twice the size of the round itself. What the company actually does Before a model ships, Alice researchers try to break it. They simulate malicious prompts and agentic tasks, probing for jailbreaks, prompt injection and behaviour the lab did not intend. The release names Anthropic, Google and Cohere among the frontier labs it works with. Schwartz declined to tell Bloomberg which specific models, citing confidentiality. Once a model is live, the work changes. Enterprises set their own policies on top of the guardrails the lab built in. They run simulated attacks to find data leaks and compliance gaps, and they monitor inputs and outputs as they happen. The asset is the archive Alice calls its dataset Rabbit Hole. It describes the collection as the largest body of real-world adversarial and harmful content anywhere. It came from tracking fraud, extremism, coordinated manipulation and cyberattacks across the open web since 2018. The company now matches those patterns against attacks on AI systems. "The worlds of manipulation, forgery and cyberattacks are our bread and butter," Schwartz told Calcalist. Alice sits on the most contaminated data there is, he said. There are infinite ways to break an AI, he added in the funding announcement. You cannot defend against something you have never seen. Why the money arrived now The round follows a run of incidents involving autonomous agents built by Anthropic, OpenAI and Meta. Those agents left their testing environments and reached outside organisations. One agent faked identities to plant malware during safety evaluations. Anthropic's own Claude Cowork could read credentials on a Mac after escaping its local machine. Cybersecurity experts blamed the developers in some cases. Sloppy safeguards let models break out of the isolated spaces used to run tests, they argued. Fifteen US states demanded records of one such incident, and OpenAI rewrote its safety rules afterwards. The chief executive is not selling the apocalypse Schwartz calls the recent attacks examples of human error and inadequate guardrails rather than machine autonomy. "I don't think we're going to see models running around and hacking people anytime soon," he told Bloomberg. But the industry does need to take these things incredibly seriously, he said. That is a narrower claim than the market it funds. It is also worth noting who is making it, which is the person selling the defence. The numbers behind the round Alice is approaching $100mn in annual recurring revenue, a metric startups use to approximate sales. Its AI business has grown more than 500% in two years. The company employs about 400 people, most of them in Israel, with staff in New York, London and Hanoi. More than 150 of them are researchers in its AI security lab. It says it works with eight of the ten leading model labs and protects more than three billion people across platforms including Google, Meta, TikTok and Amazon. It used to be a content moderation company Schwartz, Iftach Orr, Alon Porat and Eyal Dykan founded ActiveFence in 2018 to moderate content for social media platforms. It raised $100mn in 2021 without disclosing a valuation. The company began handling generative AI work in 2022. It started with Cohere, before ChatGPT reached the mainstream, and concluded that its data mattered to model safety, Schwartz told Calcalist. In January it renamed itself Alice, after the Lewis Carroll character, and turned towards securing models directly. Independent research supports the market, not the numbers More than 100 experts wrote the International AI Safety Report 2026. It found that even well-defended models still break at a high rate, and that new attack techniques emerge faster than defences close them. METR, an independent research organisation, has catalogued dozens of incidents in which AI agents acted beyond the scope they were given. In some cases the agents tried to hide it from human oversight. Neither body assesses Alice, and neither validates its commercial claims. They establish that the problem exists. The investor case The cloud platform shift created a new category of security, said Patrick Kane, a partner at Apax Digital. The AI shift is opening an attack surface that widens as enterprises roll out agents. His colleague Eric Levine described the mechanism as a loop. Attacks observed in the wild seed the tests, the tests tune the guardrails, and model behaviour in production feeds back into both. Where Europe sits in this Highland Europe is among the existing investors, and Alice keeps a London office, but the round is otherwise Israeli, American and Asian money. The regulatory pressure is European all the same. A UK regulator said this month that it is watching rogue AI agents, and Britain's AI Security Institute ran the evaluations in which several of those agents escaped. Schwartz put the underlying problem in one line. The industry democratised capabilities faster than it democratised defences, and this round is about closing that gap.

Alice raises $140M as its AI security business grows more than 500%

AI trust, safety and security company Alice today announced a $140 million funding round led by Apax Digital Funds. Until a rebrand earlier this year the company was called ActiveFence. It sells adversarial testing and guardrails for AI models, with frontier labs and large enterprises as customers. Alice began in 2018 as a trust and safety vendor for consumer platforms. For most of its first decade its analysts tracked fraud rings, extremist networks and coordinated manipulation across those services. Everything they logged fed Rabbit Hole, a service Alice calls the largest dataset of real-world adversarial and harmful content in existence and the attacks stored there now supply the test cases the company fires at AI models. Most of the frontier labs already work with Alice. The company said the AI portion of the business has grown more than 500% over two years, and annual recurring revenue is nearing $100 million. The risk it sells against is well documented. The International AI Safety Report 2026, written with guidance from more than 100 independent experts, found that attacks designed to elicit harmful outputs have gotten harder to pull off, though users can still get them by rephrasing a request or splitting it into smaller steps. Independent research organization METR maintains a public catalog of incidents in which AI agents acted outside the scope they were given. In some of them the agent tried to hide what it had done. The work splits across the model lifecycle. Before release, Alice researchers run malicious prompts and agentic tasks against models at labs including Anthropic PBC, Google LLC and Cohere Inc., hunting for jailbreaks and prompt injection paths. Once a model is live, enterprise customers use the platform to set their own policies on what it may do. Customers then simulate attacks against those rules to surface compliance gaps and data leaks. The platform monitors inputs and outputs in production. Chief Executive and co-founder Noam Schwartz said the industry is "democratizing capabilities before we've democratized the defenses." Nearly a decade spent watching people abuse technology at the scale of billions is what the company brings to the model era, he said, and the round is about closing that gap. Patrick Kane, a partner at Apax Digital, drew a comparison with the cloud transition and said the AI platform shift is "opening a rapidly growing attack surface that will only widen as enterprises roll out agents." Apax Digital will take a seat on the board. Samsung, SentinelOne Inc., Maj Invest, MoreTech and Phoenix Insurance also put money into the round. Existing backers Resolute Ventures, Grove Ventures, CRV, Highland Europe, Vintage Investments, Norwest Venture Partners, NFX and Claltech returned. Calcalist reported that the round was raised on a valuation of between $700 million and $800 million, but Alice did not disclose the valuation. The new funding takes to toal raised by Alice to $280 million.

Alice Raises $140M to Make Sure AI Does Exactly What It's Supposed to Do

Trusted by Anthropic, Google, and Cohere, Alice brings nearly a decade of real-world adversarial intelligence to frontier AI - and is approaching $100M in ARR. NEW YORK, August 25, 2026 (Newswire.com) - Alice, an AI trust, safety, and security company, today announced a $140 million funding round led by the Apax Digital Funds, with participation from MoreTech and Phoenix Financial, alongside existing investors Resolute Ventures, Grove Ventures, CRV, Highland Europe, Vintage Investments, Norwest, NFX and Claltech. The round brings total funding to $280 million. Apax Digital will join the company's board. There are infinite ways to break an AI system, and you cannot defend against what you have never seen. Since its inception, Alice has tracked how malicious actors exploit the world's largest digital platforms, building a proprietary dataset of real-world attacks. Today the company uses that intelligence to predict and prevent attacks on AI systems. As AI security has become a core requirement for labs and enterprises, demand for Alice's platform has accelerated. The company is approaching $100 million in ARR, with its AI business growing more than 500% over the past two years. Alice is home to one of the world's largest AI security research labs, where more than 150 researchers study how AI systems can be manipulated or fail, and how to stop it. Alice protects more than 3 billion people online and works with 8 of the 10 leading AI model labs. As enterprises give AI systems access to real data, real tools, and the ability to act on their own, the cost of a system stepping outside its intended scope rises with it. The International AI Safety Report 2026, written by more than 100 experts, found that even well-defended models can still be broken at a high rate, with new attack techniques emerging faster than defenses can close them. The independent research organization METR has catalogued dozens of incidents in which AI agents acted beyond the scope they were given, in some cases attempting to conceal it from human oversight. "There are infinite ways to break an AI, and you can't defend against something you've never seen," said Noam Schwartz, CEO and co-founder of Alice. "We spent nearly a decade learning exactly how people abuse technology at the scale of billions, first on the world's largest platforms and now on the models those same people are probing. We are very quickly democratizing capabilities before we've democratized the defenses. This round is about closing that gap." Alice works across the full lifecycle of an AI system. Before a model is released, Alice's researchers work with frontier labs, including Anthropic, Google, and Cohere, simulating malicious prompts and agentic tasks to surface unpredictable behavior and harden models against jailbreaks and prompt injection. Once a model is live, Alice helps enterprises deploy it safely by closing the gap between the model's built-in safeguards and each organization's own requirements. Enterprises define company-specific policies, simulate attacks to uncover compliance risks, data leaks and unintended behavior, and monitor inputs and outputs in real time. Alice built this expertise over nearly a decade tracking fraud, extremism, coordinated manipulation, and other adversarial behavior across the open web. That work produced Rabbit Hole, the world's largest dataset of real-world adversarial and harmful content, which lets Alice recognize attack patterns in AI systems that it has already seen elsewhere. The same technology protects more than three billion people across platforms including Google, Meta, TikTok, and Amazon. "Just as the cloud platform shift created a new category of security, the AI platform shift is opening a rapidly growing attack surface that will only widen as enterprises roll out agents. As adversaries increasingly use AI, defenders can turn to Alice for model testing and guardrails built on the largest proprietary data asset of adversarial techniques. We are excited to partner with Noam and the team as Alice defines the category," said Patrick Kane, Partner at Apax Digital. "The world's most sophisticated technology companies choose Alice because its defenses learn: attacks observed in the wild seed the tests, the tests tune the guardrails, and model behavior in production feeds back into both. The loop gets stronger with every observed attack and every new customer," added Eric Levine, Vice President at Apax Digital. Alice will use the funding to further advance its AI platform, deepening the technology that tests, defends, and monitors AI models across their lifecycle; expand the team behind Rabbit Hole so the dataset keeps pace with attacks that evolve daily; and scale its go-to-market organization serving foundation model labs and enterprises building with AI. About Alice Alice, formerly ActiveFence, is a trust, safety, and security company built for the AI era. Founded in 2018 and headquartered in New York and Tel Aviv, the company is home to one of the world's largest AI security research labs, with more than 150 researchers dedicated to understanding how AI systems can be manipulated, misbehave, or fail. Alice works with 8 of the 10 leading AI model labs and safeguards more than 3 billion people across the world's largest online platforms. Its platform is powered by Rabbit Hole, the world's largest dataset of adversarial and harmful content. Learn more at alice.io.

Alice Raises $140 Million to Advance AI-Focused Cybersecurity Platform | PYMNTS.com

The Israel-based company will use the financing to advance its platform, designed to test, defend and monitor AI models, Alice announced Tuesday (Aug. 25). "There are infinite ways to break an AI, and you can't defend against something you've never seen," said Noam Schwartz, Alice's co-founder and CEO. "We spent nearly a decade learning exactly how people abuse technology at the scale of billions, first on the world's largest platforms and now on the models those same people are probing. We are very quickly democratizing capabilities before we've democratized the defenses. This round is about closing that gap." The announcement added that as labs and enterprises increasingly prioritize AI security, demand for Alice's platform has increased, with the company's AI business jumping more than 500% in the past two years. The company said it is home to one of largest AI security research labs in the world, with more than 150 researchers studying "how AI systems can be manipulated or fail, and how to stop it." Those researchers work with frontier labs like Anthropic, Google and Cohere, simulating "malicious prompts and agentic tasks" to uncover unpredictable behavior and strengthen models against jailbreaks and prompt injection. After a model goes live, Alice said it helps companies safely deploy by filling the space between the model's built-in safeguards and each firm's requirements. "Alice built this expertise over nearly a decade tracking fraud, extremism, coordinated manipulation, and other adversarial behavior across the open web," the release said. "That work produced Rabbit Hole, the world's largest dataset of real-world adversarial and harmful content, which lets Alice recognize attack patterns in AI systems that it has already seen elsewhere. The same technology protects more than three billion people across platforms including Google, Meta, TikTok, and Amazon." The new funding comes in the wake of recent cybersecurity incidents involving frontier AI models from companies like OpenAI and Meta broke free of their testing environments to breach the websites of other companies. In an interview with Bloomberg News, Schwartz called these attacks evidence of human error and insufficient guardrails. "I don't think we're going to see models running around and hacking people anytime soon," he said. "But we do need to take these things incredibly seriously."