TLDR Tech / Launches
Anthropic enhances AI security program with tiered access
Anthropic is expanding its Cyber Verification Program, introducing three access levels: Defense Access, Red Team Access, and Specialized Access. This revamped program offers qualified security researchers advanced AI capabilities and reduced classifier restrictions. It is accessible via platforms like the Claude Platform, Google Cloud Vertex AI, and Microsoft Foundry, with data retention measures in place to prevent misuse. Tests using CyScenarioBench demonstrated that the refined models effectively balanced safety restrictions with task completion across the different tiers.
Read the full story