The U.S. Cybersecurity and Infrastructure Security Agency is employing Anthropic’s AI model Mythos to review government software for vulnerabilities, according to three people familiar with the effort. The initiative reflects growing federal interest in the startup’s technology amid its ongoing tensions with the White House. CISA is applying Mythos to examine code repositories for flaws that could be exploited by foreign actors or cybercriminals. Anthropic has not commented on the project. A CISA spokesperson previously indicated they would look into the matter but has not provided further details. The work is being carried out by CISA’s Attack Surface Evaluation team, which performs security assessments and simulated attacks across federal systems. Sources indicated that the reviews have identified numerous vulnerabilities, though specifics on the volume of code examined or the severity of issues remain unclear. Anthropic, which has filed confidentially for a U.S. IPO, has faced challenges in its dealings with the government. Relations deteriorated in February when the company declined to remove protections against uses involving autonomous weapons or domestic surveillance, leading to a Pentagon supply-chain risk designation that was later blocked by a court. The situation has improved following the private release of Mythos, noted for its strength in identifying cybersecurity weaknesses. The National Security Agency has reportedly used the model since April. However, the public version called Fable prompted White House demands to restrict foreign access, resulting in a temporary global shutdown that ended last week. The NSA and White House have not responded to requests for comment.
