Judge rules Trump administration illegally retaliated against Anthropic

Anthropic's Claude chatbot is one of the few AI systems cleared for use in classified settings, a position that made the company a key partner for US defense agencies. That relationship began to fracture in February, when the Pentagon demanded new contract terms allowing "any lawful use" of Anthropic's technology. The company refused, arguing its models were not reliable enough for autonomous weapons and that domestic surveillance would violate civil rights.

On Thursday, US District Judge Rita Lin ruled that the Trump administration's response to that refusal was illegal. The Department of Defense had designated Anthropic a "supply chain risk," a label historically reserved for foreign intelligence agencies and terrorists, and President Donald Trump had ordered all federal agencies to stop using the company's products. Judge Lin found the government's actions were not driven by genuine national security concerns but were retaliation for Anthropic's public stance on AI ethics.

"Citing national security is not a blank check to punish and retaliate against government critics," Lin wrote in her ruling, according to the BBC. She added that the government had failed to provide evidence of an actual supply chain risk and had bypassed legally required procedures for making that determination. The dispute centers on Anthropic's refusal to allow its AI models to be used for fully autonomous weapons or mass surveillance of Americans.

The company's CEO, Dario Amodei, made those restrictions a condition of any expanded contract with the Pentagon. Defense officials argued that private companies should not be able to constrain military action, and the White House called Anthropic "a radical left, woke company" attempting to control military activity. In that earlier order, she wrote that the challenged actions "far exceed the scope of what could reasonably address such a national security interest" and appeared "designed to punish Anthropic."

Thursday's decision confirms that the government cannot immediately enforce the blacklisting or "cripple" the company through regulatory measures, according to Reuters. Anthropic filed two separate lawsuits alleging the designation was illegal retaliation for its stance on AI safety. The company estimated the move could cost billions of dollars in lost business and reputational harm.

In a statement, Anthropic said it was "grateful the court recognized these issues need to be resolved quickly" and expressed confidence that courts would ultimately agree the supply chain designations were unlawful. The Pentagon has not yet responded to the ruling, but an appeals court in Washington, D.C. previously declined to block the blacklisting in a separate case, creating conflicting judicial signals. Further evidence in that appeals case is scheduled to be presented at a hearing on May 19.