Trump orders federal agencies to stop using Anthropic AI
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Trump announced on Truth Social Friday that he's instructing every federal agency to 'immediately cease' use of Anthropic's AI tools, with a six-month phase-out period that could leave room for further negotiations.
- Anthropic refused to amend its $200 million Pentagon contract to permit 'all lawful use' of its models, warning the change could authorize fully autonomous weapons and mass surveillance of US citizens.
- The Pentagon says it does not currently use AI for those purposes and has no plans to, but Defense Secretary Pete Hegseth met with CEO Dario Amodei earlier this week and gave the company until Friday to accept the new terms.
- The dispute escalated after Axios reported the US military used Anthropic's Claude to help plan the capture of Venezuela's Nicolás Maduro, after which a Palantir employee relayed concerns from an Anthropic staffer to military leaders.
- Sam Altman told OpenAI staff that mass surveillance and fully autonomous weapons are a 'red line,' and several hundred OpenAI and Google employees signed an open letter this week backing Anthropic and criticizing their own companies' military deals.
- Anthropic was the first major AI lab to work with the US military and remains the only one with access to classified systems, though Google, OpenAI, and xAI signed similar deals last July and built custom defense models.
Why it matters: Anthropic is the only AI company working with classified military systems, so a six-month phase-out could hand that foothold to Google, OpenAI, or xAI — the very competitors whose own employees just publicly sided with Anthropic against looser military terms. The fight is also exposing a sharp split inside Silicon Valley between executives racing into defense work and hundreds of rank-and-file engineers who consider fully autonomous weapons and mass surveillance a red line.
Ask SkimNews
