CISA uses Anthropic AI Mythos to audit government software

The U.S. Cybersecurity and Infrastructure Security Agency is using Anthropic's AI model Mythos to scan government code repositories for bugs that could leave systems open to foreign spies or cybercriminals, three people familiar with the matter told Reuters. The scans are being run by CISA's Attack Surface Evaluation team, and two sources said the audits had already uncovered a large number of vulnerabilities, though Reuters could not confirm extent or severity. The work comes amid a fraught period between Anthropic and U.S. authorities, including a Pentagon supply-chain risk designation, private agency testing of Mythos, and a recent temporary shutdown of the model's public variant.

CISA is using Mythos to scan government code for vulnerabilities