Trump Orders Agencies To Halt Anthropic Use After Pentagon Rift
President Trump has ordered federal agencies to stop using artificial intelligence products from Anthropic following a dispute involving the Pentagon, according to multiple published reports. The directive sets in motion a government-wide pullback from the company’s technology, with the Defense Department granted additional time to transition off it.
The order applies across federal departments and agencies that use Anthropic’s systems in government operations, the reports said. Several outlets described the move as a ban on Anthropic technology in government systems, paired with instructions to phase out existing usage.
The Pentagon is being treated differently under the directive. Fortune reported that the Defense Department has six months to complete its phaseout, while other agencies are directed to stop using the technology under the broader government order.
Anthropic is a major U.S. AI developer whose tools have been adopted across the private sector and, increasingly, in public-sector pilots and deployments. A government-wide halt can affect how agencies procure and test AI, how contractors support ongoing projects, and how existing tools are maintained or replaced.
The development also underscores how quickly federal AI policy can shift when national security agencies and the White House come into conflict. Even without public details of the underlying Pentagon dispute, the decision to cut off a high-profile vendor has immediate implications for the government’s approach to adopting commercial AI.
Agencies that have integrated Anthropic into workflows or systems now face the practical task of identifying where the technology is embedded and determining what must be shut down, replaced, or reconfigured. For contractors, the order can force rapid changes to deliverables, compliance requirements, and technical road maps tied to agency AI programs.
The next steps center on execution timelines and agency compliance. Departments are expected to move quickly to halt usage and begin transitions, while the Defense Department is expected to plan and complete its phaseout within the six-month window described in the reporting.
Additional details are likely to emerge through agency guidance, procurement updates, and implementation plans as departments determine how to handle existing contracts and operational dependencies. The scope of the order’s impact will hinge on how broadly Anthropic’s tools were used across federal systems and what alternatives agencies choose as replacements.
For now, the directive marks a sharp, consequential shift in the federal government’s AI posture, with agencies ordered to move off Anthropic and the Pentagon placed on a defined countdown to do the same.
