Overview
- Dario Amodei is set to meet Defense Secretary Pete Hegseth at the Pentagon on Tuesday in a decisive session over terms governing use of Anthropic’s AI.
- The Defense Department seeks “all lawful uses” of Claude, while Anthropic wants explicit prohibitions on mass surveillance of Americans and fully autonomous weapons.
- Claude is currently the only major AI model deployed on the Pentagon’s classified networks under a contract reported at roughly $200 million.
- Officials have warned Anthropic could be labeled a supply‑chain risk, a move that could void its contracts and force other contractors to certify they are not using Claude in defense workflows.
- Tensions rose after reports Claude was used in the January operation to capture Nicolás Maduro, though Anthropic says it has not discussed specific operations beyond routine technical matters; the Pentagon delegation includes Hegseth, Deputy Secretary Steve Feinberg and R&E Under Secretary Emil Michael.