Three additional cabinet-level agencies — the departments of State, Treasury, and Health and Human Services — have moved to terminate their use of Anthropic’s artificial intelligence products, widening a federal boycott that began at the Pentagon and now extends across key national security and economic institutions.
The coordinated action follows a directive from President Donald Trump ordering all federal agencies to phase out contracts with the San Francisco-based AI firm after the Defense Department labeled it a “supply-chain risk.” That designation carries significant weight in Washington, often applied to entities deemed to pose potential vulnerabilities to national security systems.
The decision represents a sharp reversal for Anthropic, whose Claude chatbot platform had been integrated into multiple government workflows. The company, backed by Alphabet’s Google and Amazon, had positioned itself as a leader in developing guardrail-heavy AI systems intended to align with democratic governance principles.
Register for Tekedia Mini-MBA edition 19 (Feb 9 – May 2, 2026).
Register for Tekedia AI in Business Masterclass.
Join Tekedia Capital Syndicate and co-invest in great global startups.
Register for Tekedia AI Lab.
On Monday, Treasury Secretary Scott Bessent announced in a post on X that the department was terminating all use of Anthropic products, including Claude. The Department of Health and Human Services notified employees in an internal message urging them to switch to alternatives such as ChatGPT and Gemini.
The State Department confirmed it was replacing the model powering its internal chatbot, StateChat, with OpenAI’s technology.
“For now, StateChat will use GPT4.1 from OpenAI,” according to a memo seen by Reuters.
State Department spokesperson Tommy Pigott said in an email: “In line with the president’s direction to cancel Anthropic contracts, we are taking immediate steps to implement the directive and bring our programs into full compliance.”
Also on Monday, William Pulte, director of the Federal Housing Finance Agency, said his bureau and mortgage finance giants Fannie Mae and Freddie Mac were ending their use of Anthropic’s products.
The directive builds on a Friday order from Trump requiring a six-month phase-out at the Defense Department and other agencies using Anthropic systems. The Pentagon’s designation of the company as a supply-chain risk escalated tensions that had been brewing over contract negotiations and the scope of AI safeguards.
At the heart of the dispute were guardrails governing military and intelligence applications. According to sources familiar with the talks, the administration and Anthropic were divided over who ultimately determines how AI systems can be deployed in sensitive defense contexts. Anthropic had pushed for firm restrictions to prevent its technology from being used for autonomous weapons targeting and domestic surveillance. The administration signaled a preference for broader operational latitude.
The fallout creates an opening for rivals, particularly OpenAI, which late Friday announced a deal to deploy its systems on the Defense Department’s classified network. OpenAI is backed by Microsoft and has emerged as a central player in federal AI procurement.
In a post on X, OpenAI Chief Executive Sam Altman said the company would “amend” its Defense Department agreement to clarify that its AI systems would not be “intentionally used for domestic surveillance of U.S. persons and nationals.” He added that the department understood the limitation to “prohibit deliberate tracking, surveillance or monitoring of U.S. persons or nationals, including through procurement or use of commercially acquired personal or identifiable information.”
The public clarification underpins the sensitivity of AI deployment in government, particularly as agencies explore applications in intelligence analysis, logistics, cybersecurity, and administrative automation. The administration’s actions signal that supply-chain integrity and alignment with executive policy priorities are now central criteria in AI vendor selection.
For Anthropic, the government-wide pullback marks a major setback. Federal contracts offer not only revenue but validation in a sector where national security credentials carry commercial weight. A supply-chain risk label from the Pentagon could complicate future bids across allied governments and defense contractors.
The broader AI industry could be impacted by the decisions. Federal procurement decisions often influence private-sector adoption, especially in regulated industries. Analysts note that by consolidating around OpenAI and other alternatives, the Trump administration is reshaping the competitive landscape at a pivotal moment when AI capabilities are rapidly advancing, and governance frameworks remain unsettled.
The situation has once again brought to the fore discussions about the AI regulatory framework. Many believed the controversy would have been avoided if there were defined rules guiding the AI industry. Now the question: who sets the boundaries of powerful AI systems — companies designing them or governments deploying them? Remains to be answered.