Six-Month Phase-Out Across Government
President Donald Trump directed all U.S. federal agencies to immediately halt the use of technology developed by artificial intelligence firm Anthropic, escalating a dispute over how the company’s models can be deployed by the military.
In a post on Truth Social, Trump said agencies currently relying on Anthropic’s products, including the Defense Department, must begin a six-month transition period to discontinue their use. He stated the federal government would no longer conduct business with the company.
Shortly after the announcement, Defense Secretary Peter Hegseth instructed the Pentagon to designate Anthropic as a “Supply-Chain Risk to National Security,” citing the firm’s refusal to comply with Defense Department demands regarding unrestricted use of its models.
Dispute Over Military Safeguards
The conflict centers on conditions attached to Anthropic’s $200 million contract with the Pentagon, signed in July. Company leadership sought assurances that its AI systems would not be used for fully autonomous weapons or mass domestic surveillance.
The Pentagon had set a 5:01 p.m. ET Friday deadline for Anthropic to agree that its technology could be used for all lawful purposes. That deadline passed without a resolution.
Hegseth stated that the company’s position was incompatible with national defense priorities, adding that Anthropic would continue servicing the department only during a limited transition window.
Industry Reaction and Broader Implications
Anthropic CEO Dario Amodei said the company could not “in good conscience” remove the proposed safeguards. He emphasized willingness to support a seamless transition if the Pentagon chose to terminate the partnership.
Meanwhile, OpenAI, which also holds a $200 million contract with the Defense Department for non-classified applications, said it maintains similar restrictions against mass surveillance and autonomous lethal weapons. CEO Sam Altman reiterated that humans should remain involved in high-stakes military decisions.
The clash underscores a widening debate between technology companies and defense agencies over ethical boundaries in AI deployment. While the Pentagon has not issued further comment, the decision marks a significant shift in the federal government’s approach to AI contractors, particularly those seeking usage limits tied to military applications.
