Anthropic to Challenge Pentagon’s ‘Supply Chain Risk’ Label in Court

Anthropic chief Dario Amodei said on Thursday the company will challenge the United States Department of War (formerly Department of Defence) in court after it labelled the AI firm a supply-chain risk, calling the move “legally unsound”.

The statement came hours after the Department formally classified Anthropic as a supply-chain risk following a weeks-long dispute over how much control the military should have over AI systems. Such a designation can bar a company from working with the Pentagon and its contractors.

The disagreement centres on how Anthropic’s AI models should be used. Amodei has said the company will not allow its technology to be used for mass surveillance of Americans or for fully autonomous weapons. The Pentagon, however, reportedly wanted unrestricted access to the systems for “all lawful purposes”.

In his statement, Amodei said the vast majority of Anthropic’s customers would not be affected by the decision.

“With respect to our customers, it plainly applies only to the use of Claude by customers as a direct part of contracts with the Department of War, not all use of Claude by customers who have such contracts,” he said.

Amodei also suggested the designation was narrower than it might appear and intended to protect government systems rather than punish a supplier.

“It exists to protect the government rather than to punish a supplier; in fact, the law requires the Secretary of War to use the least restrictive means necessary to accomplish the goal of protecting the supply chain,” he said. He added that even contractors working with the Defence Department could still use Anthropic’s AI in activities unrelated to their military contracts.

The dispute escalated after an internal memo from Amodei to staff was leaked, in which he described rival OpenAI’s work with the Pentagon as “safety theatre”, according to The Information.

After Anthropic’s fallout with the Pentagon, OpenAI has since signed a deal to work with the Department, a move that has reportedly sparked backlash among some OpenAI employees.

Amodei apologised for the leak, saying the company did not intentionally share the memo and that it was written during “a difficult day” for the company following several announcements, including a post by US President Donald Trump on Truth Social about Anthropic being removed from federal systems.