Judge Rules Trump Administration Illegally Retaliated Against Anthropic Over AI Red Lines

Federal judge sides with Anthropic in free speech dispute

A federal judge has ruled that the Trump administration illegally punished the artificial intelligence company Anthropic for publicly refusing to let the military use its technology. Judge Rita Lin found that the government's actions were retaliatory and violated the company's First Amendment rights.

The dispute began when the Pentagon asked Anthropic to remove all usage restrictions from its Claude language model. The company agreed to drop most restrictions but kept two key rules: banning mass surveillance of Americans and lethal autonomous warfare. In response, President Trump ordered all federal agencies to stop using Anthropic's technology.

Key points from the ruling

• Anthropic won on claims involving free speech, due process, and administrative law.
• The government admitted Anthropic has no backdoor access to its deployed models.
• The government acknowledged Claude is no riskier than any other AI system.
• Anthropic lost one count regarding whether Trump exceeded presidential power.

The government's case fell apart

At the center of the government's decision was a claim about national security risk. However, the government has since abandoned its central argument. Officials now concede that Anthropic does not have backdoor access to its deployed models and that Claude poses no greater risk than other "black box" AI systems.

With the security justification gone, the only remaining factor was trust. A Pentagon memo described Anthropic's behavior as having an "increasingly hostile manner through the press." Lin found the timeline of events revealed this was the true motivation.

Days before the blacklisting was announced, Defense Secretary Pete Hegseth had discussed using the Defense Production Act to make Anthropic essential to national security. The day after the designation, a government official emailed the company saying a contract was "very close." The government is still discussing collaboration on Anthropic's newest model, Mythos, across sensitive contexts.

What the judge wrote

In her 59-page order, Lin wrote: "The empty invocation of national security is not a blank check to punish and retaliate against government critics." She granted Anthropic summary judgment, meaning the facts are clear and no trial is needed.

The judge vacated the supply chain designation and issued a permanent injunction. She refused even a seven-day stay, allowing the injunction to take effect immediately.

Why this matters for AI companies

The ruling establishes an important boundary for how the government can treat technology companies that refuse certain government contracts. Officials cannot use national security as a cover to punish companies for their public positions or criticism.

This case highlights the growing tension between AI developers and government agencies over how these powerful tools should be used. As Anthropic notes, the government is still interested in working with them on future projects despite the legal dispute.

What is confirmed

• A federal judge ruled the retaliation was illegal.
• The government no longer claims Anthropic's technology is uniquely risky.
• The injunction is permanent and immediate.

What is still unclear

It remains unclear what specific damages Anthropic may seek from the government. The ruling addresses the legality of the blacklisting but does not discuss financial compensation.

• Decrypt