Secure, Safe, and Trustworthy AI Across Every App, Agent, and Model

COMPANY OVERVIEW

Alice is an AI security and safety company that helps organizations build, ship, and operate AI applications, agents, and models without fear. Founded in 2018 as ActiveFence by CEO Noam Schwartz and based in New York City and Tel Aviv, the company spent seven years collecting adversarial intelligence from the abuse of online platforms before rebranding as Alice to bring that expertise to the modern AI stack.

In August 2026 Alice raised $140 million led by Apax Digital Funds, with participation from Samsung, SentinelOne, Maj Invest, MoreTech, Phoenix Financial, and existing investors including CRV, Highland Europe, Norwest Venture Partners, and NFX, bringing total funding to $280 million. The company works with 8 of the 10 leading AI model labs, counts Google, Anthropic, Meta, Amazon, Cohere, and Nvidia among its customers, and protects more than 3 billion users across 120+ languages. Alice is SOC 2 and ISO 27001 certified.

CORE FOCUS

Alice covers the full AI lifecycle: stress-testing systems before launch, enforcing guardrails while they run, and continuously checking for drift and new risks once they are in production. The platform is aimed at foundation model labs, GenAI application teams, customer-facing agents, and regulated sectors such as financial services, healthcare, and insurance.

The differentiator is Rabbit Hole, a proprietary adversarial intelligence engine built on billions of real-world toxic and abusive samples. Because the red-teaming and guardrails draw on how attackers and bad actors actually behave rather than synthetic test sets alone, Alice can detect threats in real time across languages and modalities.

PRODUCTS & TOOLS

WonderBuild – Automated red-teaming that stress-tests AI models, apps, and agents before launch to surface vulnerabilities early.

WonderFence – Dynamic runtime guardrails that keep live AI applications safe, policy-aligned, and resistant to prompt injection and abuse.

WonderCheck – Continuous automated red-teaming in production that detects drift and emerging risks as models and threats evolve.

Rabbit Hole – The adversarial intelligence engine behind WonderSuite, powered by billions of toxic data samples and delivering real-time detection across 120+ languages.