U.S. Defense Secretary Issues Ultimatum to Anthropic: “Contract Will Be Canceled if It Does Not Agree to Military Use of AI”

U.S. Secretary of Defense Pete Hegseth has issued a final ultimatum to Anthropic CEO Dario Amodei, warning that the Pentagon could cancel its contract if the company does not accept the Department of Defense’s requirements regarding the military use of its artificial intelligence (AI) model, Claude.

Reuters and Yonhap News

On the 24th (local time), foreign media outlets including the Wall Street Journal (WSJ) and the Financial Times (FT) reported, citing sources familiar with the matter, that Anthropic must decide by the 27th whether it will agree to the terms.

On that day, Secretary Hegseth summoned CEO Amodei to Washington for a meeting. He reportedly pressured Anthropic by saying that if the company does not authorize the use of its AI model in all lawful military operations, it could be excluded from the supply chain or face the invocation of the Defense Production Act (DPA). The DPA is a Cold War-era law that allows the government, on national security grounds, to order private companies to prioritize the production and supply of certain goods.

The FT, citing a senior Pentagon official, reported that if Anthropic does not agree, the DPA could be invoked to allow the government to use the company’s technology regardless of its wishes, while simultaneously designating it as a “supply chain risk.” The WSJ also reported, quoting sources, that if Anthropic refuses, the contract could be canceled.

If the DPA is invoked, the Pentagon can use the relevant technology without a separate agreement. The law grants the executive branch the authority to prioritize the allocation of “materials, services, and facilities” for national defense. The Trump and Biden administrations used this law during the COVID-19 pandemic to address shortages of medical supplies. President Trump also applied it to expand the production of critical minerals.

If Anthropic is removed from the supply chain, significant repercussions are expected for both the company and national security-related projects, including a defense contract worth 200 million dollars. The FT also predicted that partners such as Palantir, which utilize Anthropic’s models, could be affected.

In a statement, Anthropic said, “We are engaged in good-faith discussions to support the government’s national security mission, within the bounds of maintaining the reliability and responsibility of our models.” However, if the ultimatum is carried out, the company may pursue legal action.

Meanwhile, the U.S. Department of Defense plans to enhance its military operations and threat-response capabilities through the broad use of AI technologies. In the Pentagon’s AI strategy released last month, Secretary Hegseth stated, “AI-enabled warfare and capability development will redefine the character of war over the next decade.”