- US government departments respect Trump’s anthropogenic ban
- The State Department has switched from Claude to ChatGPT
- Gemini, ChatGPT and Copilot have been approved for use in the US Senate.
The US State Department has stopped using Anthropic’s Claude model after President Donald Trump issued a directive ordering its use to immediately cease.
Anthropic fell out of favor with the US government after recently refusing to allow Claude to be used for mass domestic surveillance and autonomous weapons systems. The Pentagon received two lawsuits from Anthropic after Defense Secretary Pete Hegseth called the artificial intelligence company a “supply chain risk” and terminated the $200 million contract.
Elsewhere in government, three AI models (Gemini, ChatGPT and Copilot) were approved for use in the US Senate, with Claude notably absent.
Article continues below.
State Department migrates to OpenAI
The State Department decided to switch from Claude Sonnet 4.5 to OpenAI’s GPT-4.1 as the backbone of its internal chatbot, but the change has caused some teething problems. An internal document obtained by Nextgov/FCW claims that Claude’s move to GPT means that internal chatbot data will only be available from May 2024.
A State Department spokesperson said PakGazette“Consistent 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.”
The U.S. Department of the Treasury and the U.S. Department of Health and Human Services (HHS) are also phasing out official use of Claude, with HHS urging employees to use ChatGPT or Google’s Gemini instead.
The Senate approves the use of the big three
AI is just being implemented for Senate use, and the Senate Sergeant at Arms’ chief information officer issued a memo approving the use of Google’s Gemini chat, OpenAI’s ChatGPT, and Microsoft Copilot.
Some Senate platforms already have Copilot integrations, and data entered into Copilot is protected by the Senate’s existing Microsoft 365 secure government environment. The memo adds that Copilot can be used to “compose and edit documents, summarize information, prepare talking points and briefing materials, and conduct research and analysis.”
Details on how Gemini and ChatGPT will be used are not yet available, and questions remain unanswered about how sensitive information will be handled within the Senate’s AI ecosystem.
He New York Times notes that a House policy adopted in September 2024 prevents sensitive information from being entered into AI chatbots, but no details have been released on specific guidance.
The best cloud backup for every budget
