Refusing the Ruling: Anthropic Pushes Back Against US Government's AI 'Supply Chain Risk' Designation

Anthropic has asked the full bench of the US Court of Appeals for the D.C. Circuit to rehear a September 25 decision in which a three-judge panel ruled 2-1 in favor of the government. The company argues the majority misread the law, allowing the Department of Defense to act without any prior determination process, and that clear evidence shows the designation is retaliation for Anthropic's public statements on AI safety. The legal battle is significant. At a pivotal moment for AI, the courts must decide whether the federal government or private companies control the safety safeguards for AI technology. US developers are racing to launch the most advanced AI models and prove to investors that the technology can be profitable. Earlier this year, contract negotiations over the military's use of the company's AI broke down, leading to a public dispute with the Trump administration. The company had demanded assurances that its technology would not be used for mass surveillance of Americans or deployment in autonomous weapons. The US side said that, for national security reasons, it could not accept any restrictions on how the technology is used. In March, the Department of Defense notified Anthropic that the company and its products were designated US supply chain risks, and that all federal agencies must withdraw directives related to the company. Such risk designations have previously been applied mainly to companies from countries the US considers adversaries. Anthropic filed two lawsuits against the government: one in California challenging the Defense Department's directive for federal agencies to stop working with the company, and another in the D.C. Circuit challenging the legality of the designation itself. Anthropic won the San Francisco case. In August, US District Judge Rita F. Lin ordered the government to lift the ban on federal agencies using Anthropic's technology, stressing that the Pentagon's motive appeared to be retaliation rather than legitimate security concerns. In the Washington case, however, the three-judge panel ruled on September 25 that the government's designation was lawful. Judge Gregory Katsas, writing for the majority, said the Defense Department had reasonable grounds to worry that Anthropic might modify the design of its Claude models so that they could not perform national security tasks the department considers authorized by the contract and necessary. The case is Anthropic v. Department of Defense, No. 26-01049, before the US Court of Appeals for the D.C. Circuit in Washington.