The U.S. government filed a response in federal court on Monday denying that it unlawfully retaliated against Anthropic, while at the same time acknowledging that several agencies moved to cut off the company’s products after Anthropic resisted Pentagon demands related to military uses of its Claude chatbot.
The filing is the government’s latest reply to an action Anthropic brought on March 9, which accuses President Donald Trump and Defense Secretary Pete Hegseth of effectively blacklisting the company in retaliation for protected speech. In that March lawsuit, Anthropic asked a federal court in California to block the government from placing the company on a national security blacklist and to stop federal agencies from enforcing the designation, saying the move was unlawful and violated the company’s free speech and due process rights.
In the Monday filing, the U.S. Department of Justice also challenged Anthropic’s case on procedural grounds, contending that the ban is not subject to review because Anthropic is not challenging what the government characterizes as a 'final agency action.' The government’s procedural argument seeks to limit judicial intervention in the dispute.
An Anthropic spokesperson did not immediately respond to a request for comment on Monday.
The litigation follows the Pentagon’s formal supply-chain risk designation of Anthropic, a move that curtailed the use of a technology the filing says, and two sources previously reported, was being used in military operations in Iran. According to the court papers, Defense Secretary Pete Hegseth imposed the designation after Anthropic refused to remove safeguards it had built to prevent its AI from being used for autonomous weapons or domestic surveillance.
On March 26, U.S. District Judge Rita Lin in San Francisco temporarily blocked the Pentagon’s blacklisting of Anthropic, issuing a preliminary injunction against the designation. The legal showdown has been described as a broader test of the administration’s authority over private companies and of who ultimately controls the deployment and permitted uses of advanced AI technologies - government officials or the developers themselves.
Anthropic disclosed on June 1 that it has confidentially filed for a U.S. initial public offering. Separately, the company has a second lawsuit pending in Washington, D.C., which challenges another Pentagon supply-chain risk designation that could exclude Anthropic from civilian government contracts.
Context and implications
The government’s filing combines a denial of unlawful retaliation with a procedural challenge aimed at keeping the blacklist decision out of court for now. The dispute centers on whether the designation was lawfully imposed and whether the courts can intervene before agencies complete any further administrative steps. The litigation and the related injunction leave unresolved questions about access to government markets for AI developers and the limits of administrative power over private technology providers.