Anthropic expanded its Cyber Verification Program on Oct. 6, offering qualifying security professionals access to advanced model capabilities and reduced blocking classifiers through three application-based access tiers. The company says eligible teams can request access to Claude Opus 5.5, Sonnet 5.5, Mythos 5.1 and future models. Anthropic frames the change around the dual-use nature of cybersecurity tools: capabilities that help defenders find vulnerabilities can also assist attackers. The expansion is a gated access program, not an open release of unrestricted models. For organizations that protect public-facing services and community data, the practical question is how to extend useful testing capacity while verifying who receives access and how those tools are used.