Anthropic launches 3-tier cyber verification for AI model access

Anthropic is merging its Cyber Verification Program (CVP) and Project Glasswing into a single offering that gates access to its most capable AI models behind three tiers of cyber verification. The move signals a shift toward proactive, identity-based access control for advanced AI, potentially setting an industry precedent. Details on verification criteria and enforcement remain undisclosed.