Trump administration acted illegally against Anthropic, judge rules

A federal judge ruled the Trump administration acted illegally by designating Anthropic's AI models a "supply chain risk," finding unlawful retaliation rather than legitimate national security grounds. The decision, in the first of two Anthropic lawsuits against the government, rebukes an escalation that began with contract disputes and public attacks from President Donald Trump.

The ruling lands on August 28, 2026, as Washington and Silicon Valley continue wrestling over how artificial intelligence should serve — or stay out of — military and surveillance programs. For readers tracking how tech and politics collide, the case offers a clear snapshot of a familiar pattern updated for the AI era.

Key Takeaways

• Federal Judge Rita Lin ruled the Trump administration's supply chain risk designation against Anthropic was illegal and retaliatory.
• Anthropic filed lawsuits in March after disputes over government use of its AI for surveillance and lethal autonomous weapons.
• The designation is typically reserved for foreign companies; applying it to a domestic AI firm marked a sharp break from past practice.
• Civil liberties groups called the ruling a First Amendment win against using national security labels to punish critics.
• A second Anthropic lawsuit remains pending as negotiations between the company and federal officials continue.

What did the judge say the Trump administration did wrong?

In a 59-page ruling, Judge Rita Lin concluded that federal officials unlawfully retaliated against Anthropic despite the company's engaging in "constitutionally protected expressive activities." She wrote that "the empty invocation of national security is not a blank check to punish and retaliate against government critics."

Lin pointed to renewed conversations between Anthropic and the federal government, along with the government's decision to back away from national security allegations, as corroboration that the designation was an attempt to make a public example of the company, the New York Times reported.

Anthropic welcomed the outcome. "We welcome the court's ruling that this supply chain risk designation was unlawful," the company said in a statement. "We remain focused on working productively with the government to harness A.I. for our national security so all Americans benefit from this technology."

Why did Anthropic sue the federal government?

The legal fight erupted in March, amid a heated exchange between Anthropic, the Department of War, and President Trump over whether the government could use Anthropic's models for nationwide surveillance or lethal autonomous weapons. Anthropic's complaint accused Washington of launching a targeted campaign after contract talks broke down.

The company argued that labeling it a supply chain risk — a tool usually aimed at foreign firms posing national security threats — was an overly broad, "ideologically motivated" stretch. Trump escalated matters personally, ordering an immediate federal blacklist of Anthropic products and calling the developers "leftwing nut jobs" while branding the firm a "radical left, woke" AI company.

The public fallout was intense. Negotiations eventually resumed, but tensions did not end there. In June, Anthropic pulled its Claude Fable 5 and Claude Mythos 5 models under a federal export control directive citing security concerns about foreign nationals' access to the technology — another flashpoint in the broader dispute.

How is this fight different from how supply chain risks used to work?

Supply chain risk designations were built for a different kind of target. As Anthropic's complaint noted, the label is typically reserved for foreign companies and national security risks — not domestic AI developers locked in contract disputes with their own government.

That distinction is central to the nostalgia angle: the tools Washington once aimed at overseas adversaries now sit in a fight over Claude models, procurement policy, and political speech. Where the government once framed risk around foreign infiltration, Lin found officials here used the designation to punish a critic and send a public message.

For more stories on how yesterday's norms collide with today's headlines, explore our Nostalgia: Then & Now coverage — where legacy institutions meet modern tech speed.

What does the ruling mean for AI companies and free speech?

Civil society groups framed the decision as a warning to officials who might use security designations as leverage. Greg Nojeim, senior counsel at the Center for Democracy & Technology, said the ruling "is a vindication of the First Amendment and a clear reminder that the government isn't allowed to use national security designations to punish companies or individuals for public criticisms or business disagreements."

Nojeim added a practical line that may guide future procurement fights: "Procurement policy can't be an end run around the Constitution. The Pentagon can pick and choose which products it wants to buy. It can't use the power of its size and position to stamp out disagreement."

For AI firms negotiating defense contracts, the message is dual. Government buyers retain broad discretion over what they purchase, but punitive designations cannot substitute for lawful policy. Anthropic's partial victory in the first lawsuit does not erase the second case still working through the courts.

What happens next in the Anthropic cases?

This ruling resolves the first of two lawsuits Anthropic filed against the government. The company has signaled it wants continued engagement with federal partners even as it challenges remaining actions in court. Watch for whether export controls on Claude Fable 5 and Claude Mythos 5 face similar scrutiny, and whether other AI developers facing procurement pressure cite Lin's reasoning.

Politically, the decision gives Trump's critics a concrete judicial finding that the administration acted illegally in at least one major step against Anthropic. Whether that slows further designations — or prompts new tactics — will define the next chapter of AI governance in the United States.