AI is getting smarter. So are the people trying to break it. On August 25, AI security company Alice raises $140 million to help companies defend against attacks that didn’t exist a year ago.
The AI startup aims to solve the problem of democratizing AI capabilities without democratizing the defenses, leaving massive enterprises exposed to malicious prompt injections.
The Funding Round
Alice raised $140 million in a new funding round. Apax Digital Funds led the round. Samsung, SentinelOne, MoreTech, Phoenix Financial and others also joined, along with existing backers like CRV, Norwest and NFX.
The new money brings Alice’s total funding to $280 million. Apax Digital will join the board. Alice says it will use the cash to advance its AI platform, grow the team behind its data, and scale sales.
The company is also growing fast. Their AI business has grown to over 500% in the last two years. Revenue is now close to $100 million ARR.
More Updates: Keenable Raises $26M To Build AI Agents Search Engine
Inside the Alice Rabbit Hole
Alice is an AI trust and safety company. Formerly called ActiveFence, it’s based in New York and Tel Aviv. 3285
- Before Launch: Alice works with AI labs like Anthropic, Google, and Cohere to stress-test models. They simulate “malicious prompts and agentic tasks” to find jailbreaks and prompt injection flaws.
- After Launch: Alice provides continuous red-teaming and runtime guardrails. Companies can set their own rules, simulate breaches, and monitor AI inputs and outputs in real time.
The core is a dataset, Rabbit Hole. It was built over nearly a decade tracking fraud, extremism, and manipulation across Google, Meta, TikTok and Amazon. It’s one of the world’s largest collections of real attack data.
That history helps Alice recognize new AI attacks because it’s seen similar tricks before. Alice now protects more than 3 billion people online and works with 8 of the 10 leading AI labs. 3285
It isn’t selling to regular consumers. Their target audience is AI labs building foundation models and enterprises rolling out AI agents internally.
More Updates: Airbound Raises $37M To Scale Drone Delivery
Why This Matters Now
Noam Schwartz. Credit: Alice
CEO Noam Schwartz puts it simply: “There are infinite ways to break an AI, and you can’t defend against something you’ve never seen”. The problem is speed. AI capabilities are being released to the public very fast. But defenses aren’t keeping up.
As AI agents get more autonomy, they can do things developers never intended. Hackers are also using AI to attack, which opens a bigger attack surface.
A 2026 International AI Safety Report found defenses stay vulnerable as new attack techniques emerge. That’s why labs and enterprises are prioritizing AI security now. Alice was founded in 2018. The team spent years fighting fraud and extremism on the world’s biggest platforms.
That experience is now being applied to AI. Of all AI security research labs, the company runs one of the largest with over 150 researchers. They study the manipulations around and how to stop them.
What’s Next for Alice?
Alice will spend the funding on 3 things.
- First, deepen the tech that tests, defends, and monitors AI across their whole lifecycle.
- Second, expand the Rabbit Hole team so the dataset keeps up with attacks that change daily.
- Third, scale go-to-market to reach more AI labs and enterprises.
Apax partner Patrick Kane said it well: just like cloud created a new security category, AI is doing the same. As adversaries use AI, defenders will need Alice’s testing and guardrails.
More Updates: Standard Metrics Raises $20M for AI Portfolio Management
Click Here For The Original Source.
