Anthropic Brings Claude Mythos 5 to Claude Security: Enterprise Teams Get Frontier Vulnerability Scanning Without Direct Model Access

Anthropic has made its most advanced cybersecurity-focused model available through Claude Security, a vulnerability scanning tool for enterprise customers that connects to code repositories and identifies security flaws. The model scans code by tracing data flows across files and provides findings with severity ratings and suggested patches, but users cannot interact with the model directly through a prompt interface. This approach matters because the same capabilities that find vulnerabilities could theoretically be used to write exploits, so Anthropic restricts access to a fixed output format rather than allowing users to prompt the model freely. The company is simultaneously launching a security credit fund and expanding verification programs for its AI models' cybersecurity abilities.