Alice rebrands from ActiveFence as safety and security layer

Alice, formerly ActiveFence, rebranded after eight years of work to present itself as a foundational safety and security layer for widely used online platforms and to secure AI-powered interactions.

The company said it had protected more than 3 billion people and had worked with seven of the 10 largest Artificial Intelligence (AI) foundation models, and it named partners and clients including Amazon, TikTok, Nvidia, Cohere, and Black Forest Labs during its operating history.

Its technical approach included a proprietary adversarial intelligence engine called Rabbit Hole that continuously collected and analyzed billions of adversarial data points, citing categories such as child exploitation, fraud, prompt-injection attacks, and violent extremism, and relying on over a decade of 24/7 threat monitoring in live environments.

Alice outlined a suite of solutions organized around three protection pillars: Weaponized AI, focused on detecting and mitigating scams, deepfakes, and manipulation; Interaction Security, which applied context-aware, real-time guardrails and red teaming for user interactions; and Model Integrity, which evaluated and governed model behavior to maintain alignment with policy and compliance expectations.

“We've spent eight years learning exactly how humans can abuse tech at the scale of billions. We've tracked how harm emerges, mutates, and spreads across every major social network,” said Noam Schwartz, CEO and Co-founder of Alice.

Alice delivers end-to-end safety and security across the AI lifecycle, from model hardening and red-teaming to runtime guardrails and ongoing drift detection.