REUTERS/Dado Ruvic/ILLUSTRATION/FILE PHOTO
This image from June 5 shows the Anthropic logo, a keyboard and a robotic hand. The cyber defense agency CISA is using Anthropic’s Mythos AI model to audit government software, three people familiar with the matter said Monday.
WASHINGTON >> The cyber defense agency CISA is using Anthropic’s Mythos AI model to audit government software, three people familiar with the matter said Monday, another sign of the government’s enthusiasm for adopting the AI ​​startup’s tools even as the company is mired in an ongoing standoff with the White House.
The Cybersecurity and Infrastructure Security Agency uses Mythos to scan government code repositories for flaws that could open the door to foreign spies and cybercriminals, the sources said.
Anthropic did not respond to questions about the initiative. A CISA official said last month that he would consider whether there was anything to say on the matter, but did not respond to follow-up emails.
According to one of the sources, the scanning is carried out by CISA’s Attack Surface Evaluation team. The team is a group within CISA that conducts digital security assessments and hacking exercises across the government.
Two of the sources said the audits had already uncovered a large number of vulnerabilities, but did not elaborate. Reuters could not determine exactly how much government code the team had worked through or the nature or severity of the errors discovered.
Anthropic, which has confidentially filed for an IPO, has had a turbulent relationship with the government.
Table of Contents
Don’t miss what’s happening!
Stay up to date with the latest news, conveniently in your email inbox. It’s FREE!
Relations reached a low point in February after the San Francisco-based company refused to lift protections that prevented its AI from being used for autonomous weapons or domestic surveillance. That prompted the Pentagon to formally designate the company as a supply chain risk, a designation previously used for foreign companies suspected of facilitating espionage.
The extraordinary blacklist was blocked by a judge in March, and the conflict has eased following the private release of Anthropic’s Mythos, an AI model described as extremely capable of finding and exploiting cybersecurity vulnerabilities.
The National Security Agency, the government’s powerful eavesdropping agency, began using Mythos in April despite the blacklist, Axios reported. Late last month, The New York Times reported that NSA analysts had tested Mythos in classified environments and were impressed with its capabilities.
But when Anthropic released a public version of Mythos called Fable that included so-called cybersecurity measures, the White House suddenly demanded that foreigners be banned from operating it. This triggered a global shutdown of the model, which was only lifted last week.
The NSA and the White House did not immediately respond to requests for comment.
https://www.staradvertiser.com/2026/07/07/breaking-news/u-s-cyber-defense-agency-scans-government-software-with-anthropic-ai/
