NSA spies are said to be using Anthropic’s Mythos, although the Pentagon disputes this

The National Security Agency is reportedly using Mythos Preview, a recently announced version of Anthropic that was denied public release, Axios says. The news comes weeks after the NSA’s parent agency, the Department of Defense, called Anthropic “Supply Chain Risk,” after the company refused to give Pentagon officials access to the full range of its brand.

Anthropic announced Mythos earlier this month as a frontier model designed for cyber security applications, but said the model was too vulnerable to cyberattacks to be released publicly. As a result, the AI ​​company limited access to Mythos to about 40 organizations, which it publicly named. only twelve.

The NSA appears to be among the unspecified recipients, and is said to be using Mythos primarily for analyzing vulnerable sites. The UK’s AI Security Institute has done the same proved that it has an advantage in Mythos.

The US military’s development of Anthropic weapons comes as they say at the same time the court that those weapons could threaten the security of the country. The Pentagon conflict began when Anthropic refused to develop Claude for domestic surveillance and the development of autonomous weapons.

The NSA’s access to Mythos comes as Anthropic’s relationship with the Trump administration comes to light melting. Last Friday, Anthropic CEO Dario Amodei met with White House chief of staff Susie Wiles and Treasury Secretary Scott Bessent.

White House he says he called the meeting beneficial.

TechCrunch has reached out to the NSA for comment. Anthropic declined to comment.

Got tips or articles for the AI ​​industry? From a non-working device, contact Rebecca Bellan privately at rebecca.bellan@techcrunch.com or Signal: rebeccabellan.491.