Back to all articles
•Technology

NSA Utilizes Restricted AI Model Mythos Amidst Supply-Chain Risk Concerns

View original source

The National Security Agency (NSA) is reportedly using Mythos Preview, an advanced AI model developed by Anthropic that has been withheld from public release due to its potential for offensive cyberattacks. This revelation was reported by Axios.

  • The Department of Defense previously labeled Anthropic a "supply-chain risk" after the company denied Pentagon officials unrestricted access to the model.
  • Mythos was announced as a frontier model aimed at cybersecurity tasks. Anthropic has limited its access to around 40 select organizations, naming only a dozen publicly.
  • The NSA is purportedly using Mythos for scanning environments for vulnerabilities, with the U.K.'s AI Security Institute also confirmed to have access.
  • This development occurs as the U.S. military increases its use of Anthropic's tools, while also arguing in court over the national security threats posed by these same tools.
  • The Pentagon's disagreement with Anthropic began when the company refused to deploy its AI for mass surveillance and autonomous weapons.
  • There are signs of a warming relationship between Anthropic and the Trump administration, highlighted by a recent meeting between Anthropic CEO Dario Amodei, White House chief of staff Susie Wiles, and Secretary of the Treasury Scott Bessent.

Despite requests for comment, the NSA did not immediately respond, and Anthropic declined to comment further.