Alice, an AI security company with roots in trust and safety, has banked $140M as annual recurring revenue approaches $100M. Apax Digital Funds led the round, joined by Samsung, SentinelOne, Maj Invest, MoreTech and Phoenix Insurance, with existing backers including Norwest, CRV and Highland Europe also participating. Total funding now stands at $280M.
Formerly known as ActiveFence, Alice spent nearly a decade tracking fraud, extremism and coordinated manipulation across major internet platforms. It now applies that playbook to AI, testing frontier models before launch with malicious prompts and agentic tasks, and monitoring inputs and outputs once models run in production. The New York and Tel Aviv-based company counts Anthropic, Google and Cohere among its customers and says it works with eight of the ten leading AI model labs.
Its edge is a proprietary dataset called Rabbit Hole, built from years of abuse tracking, that underpins services safeguarding more than 3 billion people across platforms such as Google, Meta and TikTok. As companies hand agents access to databases and internal tools, the startup argues, a model producing a bad answer becomes a far smaller problem than an agent with permission to act on it.
The company plans to channel the cash into its security platform, the research operation behind Rabbit Hole and commercial expansion. Roughly 150 researchers now work on adversarial AI testing, part of a business unit that has grown more than fivefold in two years as enterprises race to deploy agents.
