Anthropic has just secured a major legal victory in its months-long fight against the Trump administration. On August 28, Judge Rita F. Lin of the US District Court for the Northern District of California ruled that the US Department of Defense’s blacklisting of Anthropic earlier this year was unconstitutional, violating its First Amendment right to free speech.
The case began last winter. When Secretary of Defense Pete Hegseth renegotiated all military contracts with AI labs, he demanded expanded rights allowing the Pentagon to deploy AI for “any lawful purpose.” Most major firms, including OpenAI, Google, and Microsoft, signed the new terms. Anthropic stood alone in maintaining its position, establishing two boundaries: prohibiting its AI from being used for mass surveillance of Americans and banning lethal autonomous weapons—systems capable of destroying targets without human intervention.
That refusal set off an escalating series of tensions. Less than 24 hours before a final ultimatum, CEO Dario Amodei issued a statement confirming the company would not waver. He wrote that Anthropic had “never objected to specific military operations or sought to arbitrarily restrict the use of technology,” but in “a narrow set of circumstances, we believe AI can undermine, rather than defend, democratic values.”
Shortly thereafter, the Department of Defense officially branded Anthropic a “supply chain risk”—a designation typically reserved for national security threats—and supplanted the company’s role by awarding contracts to seven other AI labs, including Google, Microsoft, OpenAI, and SpaceX.
Anthropic fought back with a lawsuit filed in a California court in March, alleging that the administration unlawfully retaliated against its “red lines” stance—boundaries the company refused to breach. Judge Lin sided with Anthropic early on, issuing a preliminary order restoring Anthropic’s government work while noting that the Department of Defense’s own records indicated the agency applied the supply chain risk label due to Anthropic’s “press hostility.” Punishing Anthropic for bringing the government’s contracting policies to light was “a classic case of unconstitutional retaliation,” she emphasized.
The final ruling on August 28 reaffirmed that argument. Judge Lin wrote: “A hollow invocation of national security is not a blank check to punish and retaliate against government critics.” She also characterized Secretary Hegseth’s decision as “arbitrary and capricious,” clarifying that while the Department of Defense holds full discretion to select whichever AI vendors it chooses, its sweeping punitive measures targeting Anthropic remained “unlawful and baseless.”
An Anthropic spokesperson responded that same day: “We welcome the court’s ruling that this supply chain risk designation was unlawful. We remain focused on collaborating productively with the government to harness AI for national security.”
Legally, this marks the first time an AI company has prevailed in court against the Trump administration regarding free speech in the military sphere. The decision sets a clear precedent: courts will reject the use of national security designations to penalize a corporation’s policy stance.
The dispute, however, is not over. The Department of Defense still holds the right to appeal, and broader negotiations between the government and AI labs over the military use of artificial intelligence remain an unresolved challenge. The key question ahead is whether other companies, facing comparable pressure, will dare to hold their ethical lines now that the courts have demonstrated those positions can be defended.