Newsletter Subscribe
Enter your email address below and subscribe to our newsletter
[forminator_form id="25163"]

wtvbamreuters+1reuters+1The U.S. Cybersecurity and Infrastructure Security Agency is using Anthropic's Claude Mythos AI model to scan federal government software for security vulnerabilities, Reuters reported on July 6, citing three people familiar with the matter. The deployment marks another step in the government's adoption of frontier AI for cyber defense, even as Anthropic continues to navigate a fraught relationship with the Trump administration.reuters+1
CISA's Attack Surface Evaluation team — a unit that conducts digital security assessments and hacking exercises across government — is running Mythos against federal code repositories to flag bugs that could be exploited by foreign intelligence services and cybercriminals. Two of the sources told Reuters the audits had already uncovered a large number of vulnerabilities, though none would elaborate on scope or severity.thenextweb+2
Neither CISA nor Anthropic has confirmed the work on the record. The deployment follows the NSA's testing of Mythos since April and came after the U.S. government in late June authorized Anthropic to release its Mythos 5 model to roughly 100 trusted organizations, including Fortune 500 companies and federal agencies. Commerce Secretary Howard Lutnick wrote in a letter viewed by CNBC that "appropriate safeguards are in place to permit certain trusted partners to access the Claude Mythos 5 Model".aiweekly+3
The adoption comes against a backdrop of tension between Anthropic and the federal government. In March, U.S. District Judge Rita Lin temporarily blocked the Pentagon's supply-chain risk designation of Anthropic, ruling it constituted retaliation for the company exercising its First Amendment rights. However, an appeals court in April denied Anthropic's request to extend that injunction against the Department of Defense, leaving the company barred from DOD contracts while still permitted to work with other agencies.abc7+2
Separately, Canada's Office of the Superintendent of Financial Institutions warned the country's largest banks about the cybersecurity risks posed by Claude Mythos and other frontier AI models, Reuters reported on Sunday. In an email sent in April, OSFI said "advanced artificial intelligence models, such as Anthropic Claude Mythos, significantly compress the timeframe for effective risk mitigation," urging institutions to accelerate their detection and patching capabilities.wtvbam+1
The dual-use nature of the technology — powerful enough to find thousands of critical vulnerabilities in defensive testing, yet equally capable of being weaponized by threat actors — remains the central tension as governments decide how to deploy and regulate frontier AI in cybersecurity.reuters