NSA spies are reportedly using Anthropic Mythos, despite Pentagon dispute

The NSA is said to be using Mythos Preview, a model Anthropic recently announced and which it withheld from public release. Axios reports. The news comes weeks after the Department of Defense, the parent agency of the National Security Agency, designated Anthropic as a “Supply chain risks“, after the company refused to allow Pentagon officials unfettered access to the full capabilities of its model.

Anthropic announced Mythos earlier this month as a frontier model designed for cybersecurity missions, but claimed the model was too capable of launching offensive cyberattacks to be deployed publicly. As a result, the AI ​​company has restricted access to Mythos to about 40 organizations, which it has publicly named Only twelve.

The NSA appears to be among the undisclosed recipients, and is said to be using Mythos primarily to scan environments for exploitable vulnerabilities. The UK’s AI Security Institute has also done so She confirmed that she had access To Mythos.

The US military’s expanding use of Anthropic’s tools, the company argues, comes at the same time court These tools could threaten national security. The Pentagon dispute arose when Anthropic refused to make Claude available for mass domestic surveillance and autonomous weapons development.

The NSA’s access to the Mythos comes as Anthropic’s relationship with the Trump administration unfolds Ice melting. Last Friday, Anthropic CEO Dario Amodei met with White House Chief of Staff Susie Wiles and Treasury Secretary Scott Besent.

The white house It is said He described the meeting as fruitful.

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