The Trump administration has drafted strict requirements for civilian artificial intelligence contracts that would force vendors to allow the U.S. government to use their models for any lawful purpose, the Financial Times reported.
The move comes as tensions between the Defense Department and Anthropic escalated this week. The Pentagon on Thursday designated Anthropic a “supply-chain risk,” prohibiting government contractors from using the company’s technology in work for the U.S. military. The designation follows a months-long disagreement over safeguards Anthropic sought to place on its models, which the Defense Department said were too restrictive.
A draft of guidance from the General Services Administration reviewed by the FT would require firms seeking to do business with the government to grant an irrevocable license allowing the United States to use their systems for all legal purposes. The guidance targets civilian contracts and is described as part of a wider, government-wide effort to tighten procurement of AI services, mirroring measures reportedly under consideration for military contracts.
Josh Gruenbaum, commissioner of the Federal Acquisition Service — a GSA subsidiary that helps procure software for federal agencies — told Reuters by email that continuing a business relationship with Anthropic would be “irresponsible to the American people and dangerous to our nation.” He said GSA has terminated Anthropic’s OneGov deal, removing the company from pre-negotiated contracts used across the Executive, Legislative, and Judicial branches.
The GSA draft also instructs contractors not to intentionally encode partisan or ideological judgments into AI outputs and would require disclosure if models have been modified to comply with any non-U.S. federal government or commercial regulatory frameworks. The White House did not immediately respond to requests for comment. Financial Times and Reuters reported on the developments.
