The Pentagon is deploying Anthropic’s Mythos cybersecurity model to find and patch software vulnerabilities across the US government even as it races to remove Anthropic’s products from Defense Department work in the coming months, the DOD’s top technology official said. Mythos, an unreleased Claude Mythos Preview model deployed under Anthropic’s controlled “Project Glasswing” initiative, is capable of detecting decades-old vulnerabilities in web browsers, infrastructure, and software, with DOD chief technology officer Emil Michael calling it “a national security moment” while noting Anthropic’s advantage is temporary because comparable models from OpenAI, xAI, and Google will come soon. The DOD declared Anthropic a supply-chain risk after the two sides failed to agree on how Anthropic’s models could be used by the agency, and Anthropic sued the Trump administration in March to try to reverse the Pentagon’s blacklisting.
