Sources say NSA is using Anthropic's new Mythos model despite Pentagon concerns

The scoop

It has been reported that the National Security Agency is using Anthropic's most powerful model yet, Mythos Preview. Two sources told Axios the agency is running the model now; one source allegedly said Mythos is being used more widely across the Department of Defense even though top Pentagon officials have labeled Anthropic a "supply chain risk." Anthropic CEO Dario Amodei met with White House chief of staff Susie Wiles and Treasury Secretary Scott Bessent on Friday to discuss government use of Mythos and the company's security practices.

Why it matters

This is a classic collision of urgency and unease. On one hand, the government's cybersecurity and intelligence missions crave cutting‑edge capabilities. On the other, officials worry about vendor risk and the long tail of supply‑chain exposure. Which wins out when the stakes are national security? Apparently, capability often does. The split between the Pentagon's public posture and on‑the‑ground usage underscores a deeper, awkward question: can policy keep pace with tech that moves faster than rules?

What comes next

Expect scrutiny and awkward oversight hearings. It has been reported that the breakdown in relations occurred during tense contract renegotiations earlier this year, and this episode will likely accelerate debates over procurement, vetting, and how to balance risk with operational need. For Anthropic, the high‑level meetings are a chance to ease fears — and for the government, a reminder that in the AI era, pragmatism and principle frequently find themselves in an uneasy truce.

Sources: axios.com