Anthropic to challenge DOD’s supply-chain label in court

Anthropic CEO Dario Amodei declared the company will legally challenge the Department of Defense's decision to label it a 'supply-chain risk,' a designation that can prohibit work with the Pentagon and its contractors. The dispute centers on the military's demand for unrestricted access to Claude AI for 'all lawful purposes,' which conflicts with Anthropic's public ethics policy against using its models for mass surveillance or fully autonomous weapons. The formal designation came after weeks of negotiations, which Amodei claims were productive until derailed by the leak of an internal memo where he criticized rival OpenAI's DOD dealings as 'safety theater.' OpenAI has since secured the contract in Anthropic's place.

Amodei argues the DOD's legal authority is narrow, meant to protect government supply chains rather than punish suppliers, and should apply only to direct contract work, not all usage by contractors. He apologized for the memo's tone, calling it an 'out-of-date assessment' from a 'difficult day,' and emphasized Anthropic's priority is ensuring U.S. soldiers maintain access to its AI tools during ongoing operations, which it currently supports in Iran at nominal cost. However, legal experts note challenging the designation in federal court will be difficult, as the law grants the Pentagon broad discretion on national security matters and limits standard procurement challenge avenues. The case highlights the growing tension between AI companies' ethical guardrails and government demands for unfettered technological access in defense contexts.