Federal judge rules Pentagon blacklisting of Anthropic violated First Amendment

Pentagon blacklist of Anthropic ruled unconstitutional

A federal judge in California has ruled that the Pentagon's blacklisting of Anthropic earlier this year was unconstitutional, handing the AI company a decisive win in a months-long confrontation with the Trump administration, according to The Verge. Judge Rita F. Lin of the northern district of California found that the government's actions amounted to unlawful retaliation against Anthropic for insisting on limits around military uses of its technology.

The dispute over military use limits

The conflict began this past winter, The Verge reports, when Defense Secretary Pete Hegseth moved to renegotiate the military's contracts with every major AI lab so that the Pentagon could use their systems for "any lawful use" — a significant expansion of its authority. Most labs accepted the new terms. Anthropic held out, keeping two restrictions in place: its models could not be used for mass surveillance of Americans, and they could not power lethal autonomous weapons, meaning AI systems able to select and kill targets without a human in the loop.

That refusal set off an escalating exchange, with growing pressure from both sides, public barbs from defense officials, and ultimately a final ultimatum from the administration. Less than a day before that deadline expired, Anthropic CEO Dario Amodei said the company would not budge, stating that Anthropic had "never raised objections to particular military operations nor attempted to limit use of our technology in an ad hoc manner," but that in a "narrow set of cases, we believe AI can undermine, rather than defend, democratic values."

From red lines to a supply chain risk

Soon after, per The Verge, the Pentagon labeled Anthropic a "supply chain risk" — a classification ordinarily reserved for national security threats — and worked to sideline the company within the department by signing deals with seven other AI labs, including Google, Microsoft, OpenAI and SpaceX.

Anthropic answered with a lawsuit filed in March in California district court, arguing the designation was punishment for drawing red lines. In a preliminary order, Judge Lin temporarily blocked the blacklist, pointing to department records showing Anthropic was designated because of its "hostile manner through the press," and writing that "punishing Anthropic for bringing public scrutiny to the government's contracting position is classic illegal First Amendment retaliation."

What the final ruling says

In Thursday's decision, Lin wrote that "the empty invocation of national security is not a blank check to punish and retaliate against government critics." She concluded the government's conduct was "unlawful retaliation in violation of the First Amendment" and that Hegseth's supply chain risk designation "was arbitrary and capricious." She also acknowledged limits on her finding, noting the department "is undisputedly free to select the AI vendor of its choice," while holding that "the broad measures imposed on Anthropic were illegal and baseless."

Anthropic spokesperson Danielle Ghiglieri said in a statement: "We welcome the court's ruling that this supply chain risk designation was unlawful. We remain focused on working productively with the government to harness AI for our national security so all Americans benefit from this technology."

Why it matters

The case is among the sharpest clashes yet between Washington and the AI labs it is racing to harness for defense, and the ruling draws a line around what the government can do when a vendor refuses to cooperate. At the district court level, it treats a company's publicly stated use restrictions — and its defense of them in the press — as protected expression that cannot be met with punitive contracting tools like a supply chain risk designation. For AI labs, that preserves real negotiating power: Anthropic kept its bans on mass surveillance and autonomous weapons and was vindicated rather than frozen out of federal work. For the Pentagon, the decision narrows the options available when a supplier won't bend — it may choose whom to buy from, but it cannot, in this judge's view, impose baseless designations to coerce or silence a critic. The backdrop amplifies the point: nearly every major lab accepted the "any lawful use" terms, leaving Anthropic as the outlier whose stand has now been upheld in court.