President Trump announced Friday that he is directing every federal agency in the United States government to immediately stop using technology built by Anthropic, the artificial intelligence company. Agencies like the Defense Department that currently rely on Anthropic products will get a six-month phaseout window. Everyone else? The lights go off now.
Trump made the announcement on Truth Social, and he didn't mince words:
"I am directing EVERY Federal Agency in the United States Government to IMMEDIATELY CEASE all use of Anthropic's technology. We don't need it, we don't want it, and will not do business with them again!"
Anthropic spokespeople did not immediately respond to a request for comment.
The directive arrives against the backdrop of an ongoing feud between the Pentagon and Anthropic over concerns about how the military could use AI in warfare. The details of that dispute remain murky, but the broad strokes are familiar enough: a Silicon Valley AI company apparently uncomfortable with the idea that the United States military, the institution that underwrites the very global stability these companies depend on to operate, might use its tools for national defense.
This is a pattern that conservatives have watched play out for years. Tech companies build powerful tools, sell them to the government, collect the taxpayer-funded checks, and then develop sudden moral qualms about how those tools might actually be deployed. The Pentagon isn't some rogue actor. It operates under civilian oversight, congressional authorization, and the laws of armed conflict. If an AI company doesn't want its products used by the world's most scrutinized military, it should say so upfront, before cashing the contract.
Trump's response cuts through the hand-wringing with characteristic directness. If Anthropic doesn't want to do business with the U.S. government on the government's terms, the government will find someone who will.
The AI sector has exploded in the last two years, and with it has come a gold rush of federal contracts. Agencies across the government are integrating AI into everything from logistics to intelligence analysis to administrative processing. The companies building these tools wield enormous influence over how the federal government operates, influence that, until now, has come with remarkably few strings attached.
That dynamic is shifting. When a president publicly severs ties with a major AI vendor and tells every agency to comply immediately, it sends a message that reverberates well beyond one company. Tech firms watching this unfold will take note: the federal government is a customer, not a supplicant. If you want access to the largest single buyer of technology on the planet, you play by its rules.
The six-month phaseout for the Defense Department acknowledges the practical reality that ripping out integrated systems overnight creates its own risks. That's a reasonable concession to operational continuity, not a softening of the directive's intent.
There is a deeper issue here that extends beyond any single company or contract dispute. A significant slice of the American tech industry has spent the better part of a decade signaling that it views the U.S. government, particularly the military and law enforcement, as morally suspect. Google employees revolted over Project Maven. Microsoft workers protested HoloLens contracts with the Army. Now Anthropic appears to have landed itself in a similar standoff with the Pentagon.
These companies were built in America. They benefit from American infrastructure, American capital markets, American intellectual property protections, and American security. The notion that they should then get to dictate the terms under which the American military defends the nation is not principled, it's entitled.
Trump's move treats the relationship correctly: the government sets the terms, and vendors either meet them or lose the business. That isn't heavy-handed. It's procurement.
The immediate question is implementation. Which agencies beyond the Defense Department currently use Anthropic products? What systems need to be migrated, and to which alternatives? The AI marketplace is competitive, companies like OpenAI, Google, Meta, and a growing roster of smaller firms are all vying for government work. Anthropic's loss will be someone else's gain.
The broader question is whether this moment forces the AI industry to pick a lane. You can be a company that serves the United States government and its defense apparatus, or you can be a company that doesn't. What you cannot be, what this directive makes clear you will no longer be permitted to be, is a company that takes federal money while undermining federal missions.
Anthropic built powerful technology. The government doesn't dispute that. But as Trump put it plainly: we don't need it, we don't want it. There are other vendors, other models, other partners willing to work with the country that made their industry possible.
The leverage was always on this side of the table. Now someone is using it.
