Expanding the Cyber Verification Program
We’re launching a new, expanded version of our Cyber Verification Program (CVP), which makes advanced cyber capabilities and reduced blocking classifiers available to qualifying security professionals. The program now consists of three access tiers, which allow security teams to apply for the level of access that best suits their work. Each tier includes access to our most capable models, including Claude Opus 5.5, Claude Sonnet 5.5, Claude Mythos 5.1, and new models moving forward. Interested customers can apply here.
TL;DR
- Anthropic's Cyber Verification Program (CVP) is expanded with three new access tiers: Defense Access, Red Team Access, and Specialized Access.
- Each tier offers varying levels of access to advanced AI cyber capabilities and models like Claude Opus 5.5, with tailored verification requirements.
- Defense Access is for defensive tasks like incident response and vulnerability analysis; Red Team Access includes authorized penetration testing; Specialized Access is for high-risk safety systems testing.
- Testing on CyScenarioBench showed reduced or eliminated blocks for CVP tiers compared to generally available models, with Red Team Access achieving a 67.6% success rate.
- Previous programs like Project Glasswing demonstrated significant increases in vulnerability discovery rates, with over 129,000 verified software vulnerabilities found by partners.
- Data retention is required for monitoring misuse, with Enterprise Frontier Safeguards (EFS) planned for later fall to offer zero data retention with robust safeguards.
- CVP is available on Claude Platform, Google Cloud's Vertex AI, and Microsoft Foundry, with limited availability on Amazon Bedrock for EFS-eligible customers.