Trump has banned the use of AI from Anthropic in the US public sector

US President Donald Trump has ordered all US federal agencies to "immediately stop" using the technology and software of AI startup Anthropic, he wrote in the social network Truth Social on Friday, February 27. The statement came amid pressure on AI developer Claude from the Department of Defense: the agency demanded that the company it has a contract with give it full access to AI functions.

"The left-wing lunatics at Anthropic have made a CATASTROPHIC ERROR by trying to HATE the Department of War and force it to obey their Terms of Service instead of our Constitution," Trump wrote. - Their selfishness jeopardizes the LIVES of AMERICANS, endangers our military, and undermines our NATIONAL SECURITY."

According to the president, agencies that already use Anthropic's products will be given a six-month transition period to completely phase out the company's services.

Shortly after Trump's post, Defense Secretary Peter Hegseth said on social media X that he had instructed the Pentagon to "recognize Anthropic as a national security risk in the supply chain," meaning that any company wishing to do business with the military would have to sever ties with Anthropic.

The startup signed a $200 million contract with the Pentagon in July, but sought assurances that its AI models would not be used for fully autonomous weapons or mass domestic surveillance of Americans. Anthropic's contract included working with classified material, CNBC recalls. According to Axios, Anthropic's software is already integrated into the military and Claude was used in a raid against Venezuelan leader Nicolas Maduro in January.

The Pentagon set a deadline by which Anthropic had to agree to demands that it be allowed to use the technology for all purposes - Feb. 27, 5:01 p.m. Washington time. But Anthropic CEO Dario Amodei said Thursday that the company "in good conscience cannot" authorize the use of its models without any restrictions. Hegseth called the decision "a master class in arrogance and betrayal" in a post.

Rival AI developer OpenAI also signed a $200 million contract with the Pentagon last year to use AI models in non-sensitive tasks, including routine office operations. OpenAI made it clear Friday that it is holding the same "red lines" on the use of its technology by the Pentagon or other customers as Anthropic, according to a memo from OpenAI CEO Sam Altman seen by CNBC. "We have long believed that AI should not be used for mass surveillance or autonomous lethal weapons, and that decisions of high significance should be made by humans," Altman wrote.

According to The Wall Street Journal, the head of OpenAI is working on an agreement that should resolve the standoff between Anthropic and the Pentagon and would be good for the entire industry. "We'd like to try to de-escalate," he told employees, according to the WSJ.

Separately, the 700,000-strong association of Amazon, Google, and Microsoft employees came out in support of Anthropic. In an open letter, they called on their employers to join Anthropic and refuse to comply with the Defense Department's demands for unrestricted use of AI products. Another letter was published on behalf of Google and OpenAI employees. "We call on our companies to also refuse to comply with such requirements if they, or the advanced AI labs they invest in, enter into new contracts with the Pentagon," the appeal reads.

This article was AI-translated and verified by a human editor