The Trump administration has ordered all U.S. federal agencies to immediately stop using AI company Anthropic, citing national security concerns. President Trump called Anthropic a “supply chain risk” on Truth Social, and Defense Secretary Pete Hegseth directed the Pentagon to label the company as a threat to military supply chains, effectively barring federal contracts and giving agencies six months to phase out existing Anthropic services.
The move followed months of tense negotiations over Anthropic’s AI systems, which the company sought to restrict from use in autonomous weapons and domestic surveillance. Anthropic CEO Dario Amodei has maintained that the company will not compromise on these safeguards, arguing that deploying AI for mass surveillance is incompatible with democratic values.
Hours after the Trump administration’s announcement, OpenAI CEO Sam Altman confirmed that his company had reached an agreement with the Department of Defense to deploy OpenAI models on classified military networks. Altman emphasized that the agreement upholds OpenAI’s core safety principles, including prohibitions on domestic surveillance and ensuring human oversight over high-stakes automated decisions.
The escalating conflict drew widespread attention from the AI industry. Prominent AI researchers and tech leaders signed an open letter urging the Pentagon not to penalize companies for refusing contract changes, warning that the actions could undermine U.S. leadership in AI innovation. Former Trump AI advisor Dean Ball criticized the move as “attempted corporate murder.”
Lawmakers from both parties have called for de-escalation. Senators Ed Markey (D-Mass.) and Chris Van Hollen (D-Md.) described Pentagon threats to Anthropic as a “chilling abuse of government power,” while Sen. Mark Warner (D-Va.) expressed concern that political considerations may be driving national security decisions.
Anthropic’s $200 million contract with the Pentagon for responsible AI in defense operations remains in limbo, while OpenAI will now be the primary private provider of AI models on classified networks. Altman stated that OpenAI’s approach mirrors Anthropic’s red lines on mass surveillance and autonomous weapons.
The development underscores the growing influence of private AI companies on U.S. national security and the delicate negotiations required to balance safety, innovation, and defense needs.
