A U.S. federal judge in San Francisco ruled Thursday that the Trump administration’s February sanctions against artificial intelligence company Anthropic were unconstitutional and illegal. Judge Rita Lin blocked the Pentagon’s supply-chain risk designation, finding the government unlawfully punished the creator of the Claude chatbot for public criticism.
Inside a San Francisco federal courtroom, Judge Rita Lin made clear in her 59-page ruling that executive power does not include an automatic right to muzzle corporate dissent.
Traditionally reserved for foreign entities, wielding this tool against an American AI developer sets a legal precedent for the sector.
Inside Judge Lin’s Decision on the Pentagons Blacklist
The core of the dispute traces back to a fundamental disagreement over military applications of artificial intelligence. Anthropic balked at letting the U.S. military deploy its Claude chatbot for lethal autonomous weapons and mass surveillance of American citizens. Following that refusal, Defense Secretary Pete Hegseth designated the firm a supply-chain risk, while Donald Trump publicly blasted the company as radical and out of control.
Judge Lin dismantled the administration’s legal defense. “The hollow invocation of national security is not a blank check to punish those who criticize the government,” she wrote. The court found that the sanctions were designed to make an example of a firm judged arrogant rather than address a verifiable threat. Crucially, the government dropped its own argument during the proceedings that Anthropic might sabotage its models once delivered, falling back on a loss of “trust” given the company’s public criticisms.
Before the sanctions dropped, Pete Hegseth was considering placing Anthropic under the regime of strategic industries, and the Pentagon wrote that they were “very close to an agreement” just days before the sanctions went into effect.
| Date | Event | Legal & Administrative Status |
|---|---|---|
| February | Anthropic declines military terms for autonomous weapons and surveillance. | Initial negotiations break down between the Pentagon and the AI lab. |
| February | Secretary Pete Hegseth designates Anthropic a supply-chain risk. | Sanctions implemented; Donald Trump labels the company radical. |
| March | Federal court issues provisional suspension. | Temporary relief granted as litigation begins in San Francisco. |
| Thursday | Judge Rita Lin issues 59-page ruling. | Sanctions declared illegal and unconstitutional; government retains appeal rights. |
The Global Macro-Economy and AI Sovereignty
But there is a catch. Anthropic’s victory is far from total. A separate Pentagon sanction enacted on the exact same day under the public procurement code remains in effect. A Washington court declined to suspend that measure in April, keeping the AI developer locked out of military contracts for the time being. Judge Lin’s ruling explicitly noted that the Pentagon remains free to choose alternative AI providers.
“We remain focused on a productive collaboration with the government to put AI at the service of our security nation, so that all Americans benefit from this technology,” an Anthropic spokesperson told the AFP following the verdict.
What Lies Ahead for Tech and State Power
The government retains the legal right to appeal Judge Lin’s ruling.
How do you view the balance between military AI adoption and corporate ethics? Let’s discuss this in the comments below.
