Anthropic, US government to clash over AI blacklisting
A federal appeals court in Washington, D.C., is scheduled to hear arguments on Tuesday regarding Anthropic's lawsuit challenging its blacklisting by the U.S. Department of Defense. The panel, consisting of Judges Karen Henderson, Gregory Katsas, and Neomi Rao, will hear 15-minute presentations from both the Department of Justice on behalf of the Pentagon and Anthropic. The court is expected to issue a written opinion after taking the matter under advisement. The proceedings begin at 9:30 a.m. ET. The legal battle commenced in March when Anthropic sued Defense Secretary Pete Hegseth and the DOD. The conflict arose after the DOD classified Anthropic as a supply chain risk, a designation typically reserved for foreign adversaries. This label prohibits defense contractors from using Anthropic's Claude models in their operations. The move followed a collapse in negotiations between the two sides. The DOD demanded unrestricted access to Anthropic's models for all lawful purposes, while Anthropic insisted on safeguards prohibiting the use of its technology for fully autonomous weapons or domestic mass surveillance. With no agreement reached, the DOD proceeded to blacklist the company, a decision Secretary Hegseth also criticized publicly on social media. Anthropic CEO Dario Amodei stated the company had no choice but to challenge the designation legally. While the DOD has continued to utilize Anthropic's models for operations in Iran, President Donald Trump recently noted that a deal between the startup and the military remains possible. In April, the appeals court denied Anthropic's request for a temporary stay of the blacklisting, meaning the restriction remains in effect throughout the litigation. However, the judges agreed to expedite the case, citing a likelihood of irreparable harm to the company. In its legal brief, the government argued that Anthropic poses an untenable national security risk because it could encode limitations into its future models. The briefing suggests the DOD lacks trust that the company will not manipulate its technology to enforce its own moral or policy judgments regarding military use. Conversely, Anthropic contends that the ability to encode such limits is unsupported and offers no valid basis for a risk designation. The company further argues that Hegseth and the DOD violated the Constitution and existing administrative procedures. This case is being heard in Washington alongside a separate but related lawsuit in San Francisco. Because the DOD relied on two distinct designations to justify its action, the cases are tried in different courts. In the San Francisco proceeding, Anthropic was granted a preliminary injunction allowing government agencies other than the DOD to use its models while the litigation proceeds. A judge in that case ruled that no statute supports branding an American company as a potential adversary merely for disagreeing with government policy.
