Trump orders federal government to cut ties with Anthropic

President Trump directed federal agencies to stop using Anthropic’s technology, and Defense Secretary Pete Hegseth designated the company a “supply-chain risk” — a label usually applied to firms treated as extensions of a foreign adversary. Agencies already using Anthropic’s products, including the Pentagon, were given six months to phase it out.

The dispute traced to Pentagon contract negotiations in which Anthropic refused to accept revised terms that the company said would allow two guarantees it treated as non-negotiable — that its AI would not be used to make autonomous weapons decisions without human involvement, and would not be used for mass domestic surveillance — to be “disregarded at will.” The day before, Anthropic chief executive Dario Amodei had said publicly the company had negotiated in good faith but would not compromise on those principles. Anthropic called Hegseth’s supply-chain-risk designation “legally unsound” and said it would challenge it in court, arguing the episode would “set a dangerous precedent for any American company that negotiates with the government.”

OpenAI announced its own agreement with the Pentagon the same day, embedding equivalent restrictions against autonomous-weapons use and mass domestic surveillance into its contract terms. OpenAI chief executive Sam Altman called on the Pentagon to offer identical terms to all AI companies, implicitly framing Anthropic’s exclusion as a matter of negotiating posture rather than substantive disagreement over safeguards.

The episode set two competing AI labs on opposite sides of a live confrontation between a frontier AI company and the US government over how far a vendor could bind its own product against military use — a dispute that moved into federal court within weeks as Anthropic pursued its promised legal challenge.