Trump Orders U.S. Agencies to Halt Anthropic AI, Citing Security Concerns

President Trump’s order gives federal agencies six months to purge Anthropic’s AI tools, the administration announced in a Truth Social post Friday, according to NBC News. The directive follows a Pentagon ultimatum that labeled Anthropic a “supply chain risk” if the company refused to drop its proposed guardrails on military AI use, as reported by Edition.

The dispute erupted after the Defense Department sought to rewrite a July contract with Anthropic and other vendors to remove restrictions on how AI could be deployed in combat scenarios, Wired notes. Pentagon officials pressed Anthropic to allow “all lawful U.S. military applications,” a demand the company’s CEO Dario Amodei rejected at a Davos panel on Jan. 20, according to Edition.

Trump framed the standoff as a security breach, calling Anthropic’s leadership “left‑wing nut jobs” in a Truth Social rant cited by Wired. He warned that continued use of the firm’s models would expose government systems to espionage and sabotage, echoing concerns raised by defense officials about unchecked AI capabilities.

Federal agencies have been instructed to cease all Anthropic services immediately, with a phased shutdown by the end of the six‑month window, CBS News reports. The Pentagon has not responded to requests for comment, and Anthropic’s legal team has declined to speak, per NBC News.

Analysts note that the ban could shift government AI spend toward OpenAI and Google, which have recently surged in market share, as highlighted in a VentureBeat report on AI power rankings. The move also underscores growing tensions between tech firms and the Trump administration over AI governance and national security.