A Judge Ruled the Pentagon's Blacklisting of Anthropic Was Unconstitutional
Anthropic refused to let the military use its AI for mass surveillance or killer robots. The Pentagon punished it for that. A federal judge just said that punishment was illegal.
Key points
- A federal judge in California ruled on Thursday that the Pentagon's blacklisting of Anthropic, the AI company behind the Claude chatbot, was unconstitutional.
- The Pentagon labelled Anthropic a "supply chain risk" after the company refused to allow its AI to be used for mass surveillance of Americans or lethal autonomous weapons (AI systems that can kill targets without a human making the final call).
- Judge Rita F. Lin found the designation was "unlawful retaliation in violation of the First Amendment," the constitutional right protecting free speech and the freedom to criticise the government.
- After Anthropic was blacklisted, the Pentagon signed replacement deals with seven other AI companies, including Google, Microsoft, OpenAI, and SpaceX.
- Anthropic filed its lawsuit in a California district court in March 2025 and won a temporary block on the blacklist that same month.
Anthropic walked away from a government contract fight and paid a real price for it. Then a judge stepped in.
The dispute started this past winter, when Defence Secretary Pete Hegseth moved to renegotiate all AI companies' Pentagon contracts to give the military the right to use AI for, in the department's words, "any lawful use." Most AI labs signed the new terms. Anthropic did not.
The company drew two specific lines: no using its AI to conduct mass surveillance of American citizens, and no using it to power lethal autonomous weapons, meaning AI systems that can select and kill targets without a human deciding to pull the trigger. Those were the only two restrictions Anthropic insisted on keeping.
What did the Pentagon do next?
It labelled Anthropic a national security threat. The designation it chose, "supply chain risk," is a classification normally saved for foreign adversaries or compromised hardware, not domestic companies that simply refused a contract renegotiation.
AnthropicCEO Dario Amodei published a statement less than 24 hours before a government ultimatum expired, making clear the company's position would not change. He wrote that in a "narrow set of cases, we believe AI can undermine, rather than defend, democratic values."
The Pentagon moved on, signing deals with seven other labs. Anthropic sued.
What the judge actually said
Judge Rita F. Lin did not mince words. "The empty invocation of national security is not a blank check to punish and retaliate against government critics," she wrote in Thursday's ruling. She found that the Pentagon had designated Anthropic a supply chain risk specifically because of its "hostile manner through the press," a phrase that appeared in the department's own records.
Punishing a company for going public with a policy disagreement is, the judge concluded, a textbook First Amendment violation.
Her earlier temporary order, issued in March, had already blocked the blacklist while the case proceeded. Thursday's ruling makes that conclusion permanent.
| Key moment | Date |
|---|---|
| Pentagon demands "any lawful use" contract terms | Winter 2024-25 |
| Anthropic refuses; Pentagon blacklists the company | Early 2025 |
| Amodei issues public statement refusing to budge | Before March 2025 |
| Lawsuit filed in California district court | March 2025 |
| Judge Lin issues temporary block on blacklist | March 2025 |
| Judge rules blacklist unconstitutional | June 2025 |
Anthropicspokesperson Danielle Ghiglieri said in a statement Thursday: "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."
The case was first reported by The Verge AI.
What does this mean for ordinary people?
If you use any AI tool, the principles at stake here touch you directly. This ruling says a company can set limits on how its technology is used, publish those limits publicly, and not face government punishment for doing so. That matters for anyone who cares whether AI products they rely on have safety guardrails that companies can actually defend.
For now, the court has given Anthropic its win. Whether the Pentagon appeals is the next question worth watching.
One honest takeaway: If a company's stated AI ethics policy matters to you, check whether that company actually held its position when pressure came. Anthropic did. Courts found that it paid an illegal price for doing so.
