Anthropic gives more security teams access to Claude with fewer safety restrictions

Anthropic is expanding its Cyber Verification Program (CVP) to a broader pool of security professionals, giving vetted organizations and individual researchers access to its most powerful Claude models with reduced safety filters. The access covers vulnerability research, malware analysis, incident response, and penetration testing. According to Anthropic, publicly available Claude models block most of that work because the same capabilities could aid attackers, while basic tasks such as code review or patching known flaws remain possible without special access.

The program is divided into three tiers. Defense Access is open to corporate security teams, government agencies, universities, critical infrastructure operators, open-source developers, and individual researchers. Red Team Access permits authorized attack simulations and is limited to organizations. Specialized Access is reserved for testing systems such as flight controls, power grids, or banking infrastructure, and for that tier Anthropic vets every applicant together with the US government.

Anthropic says partners in the predecessor program, Project Glasswing, found at least 129,000 confirmed vulnerabilities between April and July 2026, with more than 33,000 classified as high-severity or critical. Those figures come from surveys of a subset of partners, and Anthropic says the real-world impact is at least five times higher. Several partners reported that the AI sped up their work by months or even years.