US State Department Abandons Anthropic's Claude AI Model for Internal Chatbot
The United States Department of State has officially transitioned away from Anthropic's Claude artificial intelligence model, opting instead for OpenAI's GPT-4.1 to power its internal chatbot system known as StateChat. This significant shift follows a presidential directive from Donald Trump mandating the removal of Anthropic tools from all federal agency systems.
Internal Document Reveals Model Switch and Data Reset
According to an internal document obtained by NextGov/FCW, the State Department has specifically moved from Anthropic's Claude Sonnet 4.5 model to OpenAI's GPT-4.1. The transition has resulted in a reset of the chatbot's data, limiting it to information available as of May 2024. Previously, when operating on the Claude platform, StateChat had access to more recent training data extending to June 2025, as confirmed by an anonymous source familiar with the system.
Presidential Directive Drives Widespread AI Model Changes
The change aligns with President Trump's explicit instruction to federal agencies to cancel Anthropic contracts and eliminate their tools from government systems. A State Department spokesperson confirmed this move in a statement to NextGov/FCW, stating, "In line with the president's direction to cancel Anthropic contracts, Anthropic's Claude models are no longer available on the Department's enterprise generative AI platform. The department is taking all necessary steps to implement the directive and bring our programs into full compliance."
Migration of Custom GPT Systems and Employee Instructions
The internal document further reveals that State Department employees utilizing custom GPT systems built on Claude were instructed earlier this month to migrate to alternative government-approved models not developed by Anthropic. This comprehensive transition affects multiple AI implementations within the department, ensuring complete adherence to the presidential mandate.
Background on Claude's Government Integration and Legal Challenges
Claude had initially been introduced for federal agency use through the General Services Administration's OneGov program, which negotiated software pricing arrangements for government departments, typically for limited periods. Following the government-wide restriction on Anthropic's technology, the company has responded with legal action, filing two lawsuits this week.
The first case, filed in the US Court of Appeals for the DC Circuit, challenges the government's designation of Anthropic as a "supply chain risk" under provisions of the Federal Acquisition Supply Chain Security Act of 2018. The second lawsuit, filed in the US District Court for the Northern District of California, names several government agencies and officials as defendants and seeks injunctive relief against retaliation targeting Anthropic.
Broader Federal Agency Transition from Claude to ChatGPT
Reuters had previously reported that multiple federal agencies, including the State Department, received directives requiring them to transition from Claude to ChatGPT. This indicates a coordinated effort across the US government to standardize AI tools in response to security and policy considerations, marking a significant realignment in federal technology procurement and implementation strategies.
