President Donald Trump directed every federal agency to immediately stop using Anthropic's technology after the AI company refused to meet a Pentagon deadline on contract terms. Defense Secretary Pete Hegseth followed by designating Anthropic a national security "supply-chain risk," barring every defense contractor from using the company's AI.
Trump posted that there would be a six-month phase-out period for agencies currently relying on Anthropic's products, and clear about what comes after:
"Anthropic better get their act together, and be helpful during this phase out period, or I will use the Full Power of the Presidency to make them comply, with major civil and criminal consequences to follow."
Hegseth's statement carried the same finality:
"America's warfighters will never be held hostage by the ideological whims of Big Tech."
"This decision is final," he added. Anthropic will continue providing services for up to six months to allow for a seamless transition, but the relationship is over.
Anthropic's CEO, Dario Amodei, had made clear the company refused to drop two key demands, according to MSN reports, no fully autonomous weapons and no mass domestic surveillance. On paper, those sound reasonable. In practice, they amount to a private company dictating the terms under which the United States military can defend the nation.
Anthropic framed its position as principled. In a statement to ABC News, the company claimed:
"The contract language we received from the Department of War made virtually no progress on preventing Claude's use for mass surveillance of Americans or in fully autonomous weapons."
The company went further, alleging that new language framed as compromise "was paired with legalese that would allow those safeguards to be disregarded at will." Anthropic called the Pentagon's approach "legally unsound" and warned it would "set a dangerous precedent for any American company that negotiates with the government."
Anthropic also said it would challenge the supply-chain risk designation in court.
There is a question buried underneath the corporate PR language that matters far more than any contract clause: Does a private technology company get to define the boundaries of American national defense?
The Pentagon is not some rogue agency freelancing outside the law. Autonomous weapons restrictions and surveillance prohibitions already exist in statute and policy. What Anthropic wanted was something different: the power to impose its own conditions on top of existing law, effectively giving a Silicon Valley boardroom veto authority over military operations.
That is not a "safeguard." It is a power grab dressed in ethics language.
Consider the absurdity. A company whose entire business model depends on government contracts decided the government's own legal framework wasn't good enough. It wanted bespoke restrictions written to its specifications, enforceable on its terms. When the Pentagon set a 5 p.m. deadline to reach agreement, Anthropic let it pass, then complained it hadn't received "direct communication" on the status of negotiations.
While Anthropic was busy lawyering up, its chief competitor moved in. Late Friday, OpenAI CEO Sam Altman announced a deal with the Pentagon to deploy its AI models on the department's classified network.
Altman struck a notably different tone from Amodei:
"Two of our most important safety principles are prohibitions on domestic mass surveillance and human responsibility for the use of force, including for autonomous weapon systems."
He said the Pentagon "agrees with these principles, reflects them in law and policy, and we put them into our agreement." Same principles. No standoff. No threats of litigation. No public drama. Altman even called on the Pentagon to offer the same terms to every AI company, adding that he wanted to see things "de-escalate away from legal and governmental actions and towards reasonable agreements."
The contrast is instructive. Two companies, both operating in the same space, both holding similar safety commitments. One found a way to work with the Defense Department. The other decided to grandstand.
Adam Conner of American Progress wasted no time casting this as government overreach, writing that the Pentagon "could essentially declare war not on a foreign nation but on one of America's most successful frontier AI companies if it does not bow to its demands." He called it "an unprecedented and unnecessary peacetime move" meant to signal that companies "must do the Trump administration's bidding or face existential consequences."
This framing reveals more about the left's priors than about the actual dispute. The federal government choosing not to do business with a company that refuses its terms is not "declaring war." It is procurement. Every government contract comes with conditions. Every contractor is free to walk away. And every administration is free to find a vendor that will actually deliver.
The notion that a company should be able to reject the government's requirements and still keep the government's money is a concept that exists nowhere outside progressive think tanks.
Members of the Senate Armed Services Committee sent a private letter to both Hegseth and Amodei ahead of the deadline, urging them to resolve the dispute and extend negotiations. The letter acknowledged that the Pentagon "does not intend to conduct mass surveillance or use autonomous weapons without humans on the loop," positions the senators endorsed.
But the senators also conceded a telling point:
"It is clear, however, that the issue of 'lawful use' requires additional work by all stakeholders. We must determine whether additional legislative or regulatory language is required, and, if so, what that law and regulation should entail."
That is Congress's job. Not Anthropic's. If lawmakers believe current statutes need updating, they can write new ones. What they should not do is encourage a private company to freelance its own regulatory framework and impose it on the military through contract negotiations.
Anthropic now faces a six-month clock, a supply-chain risk designation, a ban from defense contracting, and expulsion from every federal agency. It says it will fight in court. That is its right.
But the company should understand what it chose. It chose ideology over access. It chose a public fight with the Pentagon over a quiet deal that its competitor signed the same day. It chose to let a Silicon Valley worldview override the judgment of the people Americans elected to make defense decisions.
OpenAI read the room. Anthropic read its own press releases.
The government of the United States does not negotiate at the pleasure of its vendors. Anthropic just learned that the hard way.