DC appeals court sides with Pentagon on blacklist of Anthropic

The U.S. Court of Appeals for the D.C. Circuit sided with the Trump administration in its blacklisting of the artificial intelligence firm Anthropic, allowing the Pentagon to label the AI firm a supply chain risk.

The 2-1 ruling, issued Friday, rejected Anthropic’s challenge of the designation, finding the Department of Defense had “ample support” for its belief that the continued use of Anthropic’s Claude models by the agency or its contractors presented a national security risk.

The three-judge panel in D.C. specifically took up the question of whether the Pentagon and Secretary Pete Hegseth had the legal authority under a statute of the Federal Acquisition Supply Chain Security Act (FASCA) to declare Anthropic a supply chain risk.

“The Department reasonably feared that Anthropic might manipulate Claude’s design to prevent it from performing national-security functions that the Department deems contractually authorized and necessary,” Circuit Judge Gregory Kastas wrote for the majority.

Anthropic sued the Trump administration in March, challenging both the Pentagon’s designation — typically reserved for foreign adversaries — and President Trump’s directive for civilian agencies to stop using Anthropic’s products after negotiations fell apart over safety guardrails earlier this year.

Anthropic requested its technology not be used in fully autonomous lethal weapons or for the mass surveillance of Americans, while the Pentagon insisted it be allowed to use Anthropic’s Claude for “all lawful uses.”

“Because Anthropic was willing and able to enforce contractual restrictions through model training, the Department reasonably worried that ‘critical defense system’ supported by Claude might ‘fail to engage’ as the Department would expect,” Kastas wrote.

Kastas pointed to Anthropic’s acknowledgement that the company includes certain safety guardrails that can prevent certain uses of the models.

The company’s usage policy states its normal models cannot be used for disinformation campaigns, malicious cyber operations, censorship and domestic surveillance.

But it also offers Claude Gov models, which have fewer refusals when dealing with classified information.

“On more than one occasion, these restrictions have stopped Claude from performing tasks requested by government users,” Kastas continued.

Kastas referred to an incident disclosed by the government in which staff at the Centers for Disease Control and Prevention (CDC) used a commercial Claude model in 2025 and some prompts were refused.

The court further determined Anthropic’s constitutional claims related to the First and Fifth amendments were unfounded.

“Its due-process claim fails because the Department promptly notified the company of the exclusion and its supporting rationale, and then gave the company a fair opportunity to contest the exclusion,” Kastas wrote.

“And Anthropic’s First Amendment, claim fails because the Department excluded Anthropic from its supply chain based on the company’s refusal to assent to a contract term that the Department deemed essential, not based on the company’s support for greater governmental regulation of AI technology,” he continued.

Judge Neomi Rao joined Kastas in the majority, while Judge Karen LeCraft Henderson dissented, arguing the government interpreted the supply chain risk statute too broadly.

Friday’s ruling breaks with a California federal judge’s decision last month finding the government violated the First Amendment when it issued the supply chain risk.

Emil Michael, the Pentagon’s under secretary for research and engineering who was central to negotiations with Anthropic, celebrated the ruling Friday.

“The hammer of justice has smashed [Anthropic’s] arguments. They are a Supply Chain Risk to the defense industrial base serving the [Department of War]”, Michael wrote on X. “Warfighters will sleep better knowing that no private company will insert their opinions in the chain of command. [Secretary Hegseth] was right!”

Federal judge Rita Lin agreed with Anthropic’s argument that the Trump administration retaliated against the Claude maker for its “constitutionally protected activity” and violated the due process clause of the Fifth Amendment.

The contradicting rulings could tee up a Supreme Court fight in the future.

A spokesperson for Anthropic said the firm “respectfully disagrees” with the appeals court ruling.

“Another federal court has already held the government’s parallel designation unlawful,” they added. “We remain confident in our position and are considering all options, including further review.”