Negotiations between the Pentagon and AI company Anthropic became more complicated on Friday when President Donald Trump announced on Truth Social that the government would stop using the company’s technology. The president ordered the Pentagon to begin a six-month program, accusing Anthropic of being run by “Leftwing nut jobs.”
The Truth Social newspaper said that Anthropic wants the government to follow its working principles.
“THE UNITED STATES OF AMERICA WILL NOT ALLOW A RADICAL LEFT CAMPAIGN, AWAKENING TO HOW OUR GREAT ARMIES FIGHT AND WIN WARS!,” Trump wrote. “That decision belongs to YOUR COMMANDER-IN-CHIEF, and the great leaders I have appointed to run our Army. The Leftwing nut jobs at Anthropic have made a HUGE MISTAKE by trying to STRENGTHEN the Department of the Army, and forcing them to obey their Terms of Service instead of our Constitution.”
Credit: Screenshot courtesy of Truth Social
The federal government and Anthropic have been at odds for weeks as they try to come to an agreement on how the military can use Claude, Anthropic’s AI model. Anthropic CEO Dario Amodei has been adamant that he will not allow the Pentagon to use Claude to spy on Americans or create autonomous weapons, such as unmanned drones.
The government reportedly agrees to those terms, according to the report New York Timesbut the legal language of the contract provided a lot of wiggle room for Anthropic’s relief. Anthropic is known for taking a cautious approach to AI development, and its founders left OpenAI due to AI security concerns.
On Thursday, Amodei explained his situation in a blog post:
Mashable Light Speed
“Anthropic understands that the Department of Defense, not private companies, makes military decisions. We have never opposed any military operation or tried to limit the use of our technology in a temporary manner.
However, in a small set of cases, we believe that AI can undermine, rather than protect, democratic values. Some uses are also outside the bounds of what today’s technology can do safely and reliably.”
A Friday evening deadline was set for an agreement between the Pentagon and Anthropic. It’s unclear whether Trump’s announcement of the phase-out will amount to more time for negotiations or if the government will move forward with ousting Anthropic by declaring it a supply chain risk. The government may also seek to force Anthropic to agree to its terms of the Defense Production Act, according to Times. The government may also choose another AI partner, such as Elon Musk’s Grok, but CIA officials believe the product is inferior to Anthropic, Times reports.
After the president’s announcement Friday afternoon, OpenAI CEO Sam Altman appeared on CNBC and voiced his support for Anthropic. “For all the differences I have with Anthropic, I have a lot of faith in them as a company and I think they care about safety, and I’m glad they’ve been supporting our battles,” Altman said, according to a clip posted to X.
This Tweet is currently unavailable. Either it is loading or it has already been downloaded.
Meanwhile, dozens of employees from Google and OpenAI, both Anthropic competitors, signed letters supporting Amodei’s stance. And outside Anthropic’s San Francisco headquarters, words of support appeared in chalk on the sidewalk, according to a post on X.
This Tweet is currently unavailable. Either it is loading or it has already been downloaded.
This week, Anthropic is softening its security policy — often considered one of the strictest in Silicon Valley — citing the reluctance of competitors to follow suit and the federal government’s lack of prioritization of security.
“The policy landscape has shifted towards prioritizing AI competition and economic growth, while security-focused discussions have yet to achieve meaningful results at the government level,” the company wrote. “We remain convinced that effective government engagement on AI security is necessary and achievable, and we intend to continue to advance the discussion based on evidence, the interests of national security, economic competitiveness, and public trust. But this proves to be a long-term project – not something that happens naturally as AI becomes more powerful or exceeds certain limits.”
Disclosure: Ziff Davis, Mashable’s parent company, in April 2025 filed a lawsuit against OpenAI, alleging that it infringes Ziff Davis’s copyright in training and using its AI programs.
