Trump orders federal agencies to halt Anthropic AI use
Get the Geopolitics newsletter
Daily geopolitics — wars, elections, sanctions, the diplomatic moves that move markets. Free.
- Trump directed every federal agency to immediately cease using Anthropic's technology, granting a six-month phaseout for the Department of Defense and other agencies whose systems already embed the company's AI.
- Anthropic holds a $200 million contract with the Pentagon, and CEO Dario Amodei said his company "cannot in good conscience accede" to the Defense Department's demands for unrestricted military use of Claude.
- The directive landed just over an hour before the Pentagon's compliance deadline, after the Pentagon threatened to invoke the Defense Production Act against Anthropic and considered designating the lab a supply-chain risk.
- The standoff centers on Anthropic's demand for narrow assurances that Claude won't be used for mass surveillance of Americans or in fully autonomous weapons — red lines Amodei said were undermined by contract "legalese" that would let the safeguards be "disregarded at will."
- OpenAI CEO Sam Altman publicly sided with rival Anthropic, calling the Pentagon's posture "threatening" and telling CNBC the broader AI field shares Anthropic's red lines, a notable break given that Amodei co-founded OpenAI before leaving in 2021.
- Senator Mark Warner (D-VA), vice chairman of the Senate Intelligence Committee, said the directive "raises serious concerns about whether national security decisions are being driven by careful analysis or political considerations."
Why it matters: A $200 million Pentagon contract is being torn up over red lines (no autonomous weapons, no mass surveillance) that Anthropic's biggest commercial rival, OpenAI CEO Sam Altman, publicly endorses. The six-month phaseout forces a military migration off Claude already embedded in intelligence and armed-services systems, while giving Anthropic's competitors an opening to capture the national-security business. Senator Warner's warning that politics, not analysis, is driving the decision points to a chilling precedent: AI vendors that draw ethical lines with government buyers may now face retaliation.
Ask SkimNews
