Alice has raised $140 million in a new funding round. The deal values the company at about $800 million, close to unicorn status, and brings its total funding to $280 million. Apax Digital Funds led the round, which was announced on August 25, 2026.
Alice is an AI trust, safety, and security company. Until a rebrand earlier this year, it was known as ActiveFence. It tests AI models for weaknesses and builds guardrails that keep them in line once they are live. This matters more as companies roll out AI agents, which can act on their own and be tricked into doing things they should not. You can read about the wider trend in our piece on the AI agent security funding surge.
The round at a glance
| Detail | Info |
|---|---|
| Amount | $140 million |
| Valuation | ~$800 million (near $1 billion) |
| Date | August 25, 2026 |
| Lead investor | Apax Digital Funds |
| Total raised | $280 million |
| Revenue | Approaching $100M ARR |
New investors Samsung, SentinelOne, Maj Invest, MoreTech, and Phoenix Insurance joined the round. Existing backers Resolute Ventures, Grove Ventures, CRV, Highland Europe, Norwest, and NFX also took part. Apax Digital will join the company’s board.
What Alice does
Alice spent nearly a decade as ActiveFence, tracking how bad actors abuse large internet platforms. That work built a huge dataset of real-world attacks, which the company calls Rabbit Hole. Alice now uses that data to attack AI models on purpose, a process known as red-teaming, to find where they break.
It then turns those findings into guardrails that run in production and stop a model from stepping out of line. The loop feeds itself. Attacks seen in the wild become new tests, the tests tune the guardrails, and live model behavior feeds back into both.
The company was founded in 2018 and is based in New York and Tel Aviv. Its CEO and co-founder is Noam Schwartz.
The growth behind the round
Alice is growing fast. It says it is approaching $100 million in annual recurring revenue, after growing more than 500 percent over the past two years. It reports working with eight of the top ten AI model labs, and counts Anthropic, Google, and Cohere among the companies that trust it.
Why it matters
AI security is becoming its own budget line. As companies put models into products and give agents access to real data and real tools, the risk of a system acting out of scope grows with it. A 2026 international safety report found that even well-defended models can still be broken at a high rate, with new attack methods arriving faster than defenses can close them.
That is the shift behind this round. Alice turned AI safety into a commercial product with real recurring revenue, rather than a research project. It joins a wave of security startups raising large sums to protect AI systems and the agents built on them.
What is next
Alice plans to use the money to build out its security platform, grow the research team behind its Rabbit Hole dataset, and expand its commercial operations. The bet is that model testing and guardrails become a standard purchase for every company deploying AI at scale.
Frequently asked questions
Alice raised $140 million in a round led by Apax Digital Funds, announced on August 25, 2026. The deal brings its total funding to $280 million.
Alice, formerly known as ActiveFence, tests AI models for weaknesses through red-teaming and builds guardrails that keep models and agents from acting out of scope once they are live.
The round valued Alice at about $800 million, close to a $1 billion valuation.
Apax Digital Funds led the round, with new investors including Samsung, SentinelOne, and Maj Invest, alongside existing backers like CRV, Highland Europe, Norwest, and NFX.
As enterprises deploy AI models and agents that can act on their own, the risk of misuse grows. Companies are treating model testing and guardrails as a mandatory cost, which is driving fast growth and large rounds in AI security.