Select Language:
On Monday, three additional U.S. cabinet agencies—the Departments of State, Treasury, and Health and Human Services—announced plans to stop using Anthropic’s AI products, following the Pentagon’s lead under a new White House order. This move signals a broad industry shift away from Anthropic and its chatbot platform, Claude, representing a challenge to a key player that had positioned the U.S. as a leader in security-sensitive AI.
President Donald Trump ordered all federal agencies to cease using Anthropic’s technology, citing it as a potential supply chain risk identified by the Defense Department. This designation could severely limit Anthropic’s participation in government contracting, effectively branding it as a problematic supplier.
In alignment, Treasury Secretary Scott Bessent revealed viaX that his department would cease using all Anthropic products, including Claude. The Department of Health and Human Services also instructed its employees to consider alternative AI platforms, such as ChatGPT and Google’s Gemini, though it did not immediately comment on the specifics.
Similarly, the State Department announced it would switch its in-house chatbot, StateChat, to OpenAI’s technology, moving away from Anthropic. An internal memo indicated they would employ GPT-4.1 for now, with more details to follow. State Department spokesperson Tommy Pigott confirmed the department was acting swiftly to implement the new directives.
William Pulte, head of the Federal Housing Finance Agency, also declared that his agency—and Fannie Mae and Freddie Mac—would revoke all Anthropic-based services. These developments come after Trump’s earlier order on Friday to phase out Anthropic’s products within six months across defense and other agencies. Anthropic’s backers include Google’s parent company Alphabet and Amazon, highlighting its significant industry ties.
The actions come amid ongoing disagreements over safeguards to prevent military or intelligence misuse of AI—specifically, concerns about autonomous weapons and domestic surveillance. Negotiations between Anthropic and military agencies reportedly stalled over these safety measures.
Meanwhile, OpenAI, supported by Microsoft and Amazon, announced on Friday that it secured a new agreement to deploy its AI within the Department of Defense’s classified network. CEO Sam Altman stated on Monday that OpenAI plans to amend the arrangement to clarify that its systems will not be used explicitly for domestic surveillance of U.S. citizens. He emphasized that the department is aware of and agrees to restrictions on deliberate tracking and monitoring.
ChatGPT
Perplexity
Gemini AI
Grok AI
