Anthropic’s cybersecurity-focused AI model is reportedly helping U.S. officials identify software vulnerabilities across government systems, marking a notable shift after the company weathered a contentious dispute with the Donald Trump administration.
CISA Reportedly Uses Mythos AI To Scan Government Code
The U.S. Cybersecurity and Infrastructure Security Agency, or CISA, is using Anthropic’s artificial intelligence model, Mythos, to audit government software for security flaws, Reuters reported Monday, citing three people familiar with the matter.
According to the report, CISA’s Attack Surface Evaluation team is deploying the model to analyze government code repositories for vulnerabilities that could potentially be exploited by foreign intelligence agencies or cybercriminals.
Two of the sources told the publication that the AI-powered audits have already uncovered a significant number of software vulnerabilities.
