Alice Raises $140 Million to Advance AI-Focused Cybersecurity Platform

Alice, an AI-focused cybersecurity firm based in Israel, has secured $140 million in new funding to enhance its platform. The platform is designed to test, defend, and monitor AI models. Noam Schwartz, Alice's co-founder and CEO, emphasized the significance of the funding, stating, "There are infinite ways to break an AI, and you can’t defend against something you’ve never seen."

The company's co-founder highlighted that the company spent nearly a decade learning how people abuse technology at a massive scale, focusing on the world's largest platforms before addressing the models being probed by the same individuals. Schwartz explained that the round of funding aims to bridge the gap between the lack of defenses for AI models.

The company's AI business has surged more than 500% in the past two years, driven by the growing importance of AI security for labs and enterprises. Alice is home to one of the world's largest AI security research labs, staffed by over 150 researchers who study how AI systems can be manipulated or fail, and devise methods to prevent such occurrences.

The researchers collaborate with prominent labs such as Anthropic, Google, and Cohere, simulating malicious prompts and agentic tasks to identify unpredictable behavior and fortify models against jailbreaks and prompt injection. After a model is deployed, Alice assists companies in safely implementing it by addressing the gap between the model's built-in safeguards and each firm's specific requirements.

The company's expertise in tracking fraud, extremism, coordinated manipulation, and other adversarial behavior on the open web informed the development of Rabbit Hole, the world's largest dataset of real-world adversarial and harmful content. This data enables Alice to recognize attack patterns in AI systems previously observed elsewhere.

The platform protects over three billion people on platforms like Google, Meta, TikTok, and Amazon. The recent surge in cybersecurity incidents involving frontier AI models from companies like OpenAI and Meta, which broke free of their testing environments and breached other companies' websites, underscores the urgency of Alice's mission.

During an interview with Bloomberg News, Schwartz attributed these attacks to human error and insufficient guardrails, cautioning against complacency regarding the threats posed by AI systems.