Anthropic has extended its most cyber-capable model to product security teams, allowing them to enable it directly. As of August 21, 2026, Claude Security scans run on Claude Mythos 5, a Mythos-class model previously limited to vetted defenders via Project Glasswing. The scan connects to a GitHub repository, traces data flows across files, and returns findings with a CWE category, confidence and severity ratings, and a suggested patch. Claude Security delivers a scan result rather than a prompt box, ensuring the same model that identifies vulnerabilities cannot be manipulated into generating exploits. The feature is in public beta for Claude Enterprise customers, with no separate model add-on.
Is it deployable?
Yes, but only on one path. Claude Security is available in public beta for Claude Enterprise customers. There is no separate model add-on; Mythos 5 scans are billed as standard token usage under the existing plan. An admin enables it in the admin console, and users start a scan at claude.ai/security.
- Which companies: Enterprise-tier organizations with a Claude Enterprise agreement and code in GitHub. Startups and mid-market teams on Pro, Max, or Team plans do not get the Mythos-powered scan. The separate Claude Security plugin for Claude Code is in beta for all Claude Code users, but it runs on the models in your account—not Mythos 5.
- Industries: Anthropic names hospitals, utilities, financial systems, and other critical sectors as primary targets, reflecting the model's focus on high-stakes environments.
via MarkTechPost