Anthropic CEO says talks continue with Pentagon after risk label

In short: Anthropic CEO Dario Amodei says the company is still in talks with the Pentagon after the US Defense Department labeled Anthropic a “supply-chain risk” during a dispute about limits on how its AI can be used.

What happened

Anthropic, the company behind the Claude AI models, signed a contract with the Department of Defense in July 2025 to work on AI technology. The relationship turned public and tense on February 26, 2026.

That day, Amodei said Anthropic would not remove certain restrictions. These restrictions block its AI from being used in autonomous weapons (weapons that can act on their own) or mass surveillance (watching large groups of people). Hours later, Defense Secretary Pete Hegseth posted on X that Anthropic was being designated a “supply-chain risk” to national security, which would push military contractors to stop using Anthropic products after a six-month transition period.

Anthropic later said it had not received formal notice and called the label retaliatory. The company said it would challenge any formal action in court, while also offering to help with a smooth handover during the transition.

On March 3, Amodei said at a Morgan Stanley conference that Anthropic is still discussing the issue with the Pentagon, aiming to “de-escalate” and find terms both sides can accept. So far, there has been no confirmed follow-up action from the Department of Defense beyond the public announcement.

Why it matters

This dispute is about who gets to set the rules for powerful AI tools used in defense. It also affects competition, since rivals like OpenAI have announced their own Defense Department deal, potentially taking Anthropic’s place if the split becomes permanent.

Source: The Verge AI