HHS Tells Employees to Stop Using Anthropic’s Claude

The Department of Health and Human Services is disabling its employees’ access to the AI platform Claude after President Donald Trump declared its parent company, Anthropic, a “supply chain risk.”

“As of today, users will no longer be able to log in to or access Claude through the HHS enterprise environment,” read an email sent to HHS employees on Monday from the Office of the Chief Information Officer and reviewed by NOTUS. “Please discontinue any ongoing use of the platform and transition your work to one of the Department’s other approved enterprise AI solutions.”

The order comes after Anthropic and the Department of Defense failed to reach a deal allowing the agency unfettered access to use the company’s systems as it saw fit. Trump wrote on Truth Social that “The Leftwing nut jobs at Anthropic have made a DISASTROUS MISTAKE trying to STRONG-ARM the Department of War, and force them to obey their Terms of Service instead of our Constitution.”

Anthropic’s CEO, Dario Amodei, said in a statement that the company’s decision was due to concerns over its technology being used for mass domestic surveillance and fully autonomous weapons.

The email sent to HHS employees noted that ChatGPT and Google Gemini remain available to federal workers. But one Food and Drug Administration employee told NOTUS that of all the AI tools made available to them as the FDA pushes for greater integration of the technology into its practices, Anthropic’s Claude Sonnet 4.5 is by far the most useful.

“For serious scientific work, study reviews and regulatory writing or R-coding, there is no substitute,” the FDA employee, who requested anonymity to speak candidly about their employer, said. “For an entire year we have been training reviewers on writing prompts and using the [large language model], and finally it looks like the AI was catching on. This decision, if confirmed, will basically wipe out 18 months of efforts.”

Under the Trump administration, the FDA’s AI initiatives have accelerated in an effort to speed up the time it takes agencies to review applications from companies seeking FDA approval for their products. The agency released its own AI platform, Elsa, last year, giving employees access to Claude and other AI tools.

Trump’s Truth Social post called for a six-month phase-out period for agencies currently using Anthropic’s tools. The FDA employee said that they still had access to Claude through Elsa just after receiving the email telling employees to transition to other AI platforms. But a banner on an internal FDA website said Elsa would be transitioning away from Anthropic models.

“Gemini is already available in Elsa and will become the primary model going forward,” the banner read. “While both models remain available at this time, we encourage users to begin exploring Gemini as we prepare for this change.”

But the FDA employee told NOTUS they were skeptical about whether Google Gemini could effectively replace Anthropic’s AI tools.

“Gemini is great as a chatbot to mimic a human in social media interactions, but I am not looking for an AI to flirt with,” the employee said. “I am looking at it doing in-depth analysis of complex scientific study reports.”