Anthropic Expands Cyber Access Into Three Tiers

Anthropic has expanded its Cyber Verification Program into three tiers, combining it with Project Glasswing to give vetted security organizations different levels of access to advanced Claude models for cybersecurity work. The company says the program is intended to support defense and authorized testing while retaining safeguards against harmful uses. Announced October 6, the expanded program offers Defense Access, Red Team Access and Specialized Access. The tiers include Claude Opus 5.5, Claude Sonnet 5.5 and Claude Mythos 5.1, with future models also eligible for inclusion, according to Anthropic. Applicants must undergo verification and meet security-control requirements tailored to the work they propose. Defense Access covers tasks such as incident response, malware analysis and vulnerability assessment. Anthropic says security teams at businesses, nonprofits, universities and government organizations may qualify if they protect systems they own or maintain. Critical infrastructure operators, open-source maintainers, smaller security firms and individual researchers with a record of reporting vulnerabilities are also among potential applicants.