War Secretary Pete Hegseth is close to cutting business ties with Anthropic and designating the firm a “supply chain risk,” a move that would require anyone seeking to do business with the U.S. Department of War to sever ties with the company.
Chief Pentagon spokesman Sean Parnell said the Department of War’s relationship with Anthropic is under review, stating partners must be willing to help warfighters win and protect the safety of the American people. Anthropic’s Claude is the only AI model currently available in the military’s classified systems and has been widely used, including during the January raid targeting Nicolás Maduro.
Months of negotiations between Anthropic and the Pentagon have grown contentious over how the military can use Claude. Anthropic is willing to loosen terms but wants safeguards preventing mass surveillance of Americans and the development of weapons that fire with no human involvement, while the Pentagon says those limits are too restrictive, with gray areas that make such terms unworkable, insisting AI tools be available for “all lawful purposes.”
The contract under threat is valued at up to $200 million. The dispute is shaping negotiations with OpenAI, Google, and xAI, which have agreed to remove safeguards for military use in unclassified systems but are not yet deployed for classified work.
