Court Upholds Pentagon's Anthropic Blacklist in 2-1 Ruling

A federal appeals court on Friday upheld the Pentagon's blacklisting of Anthropic from military contracts, ruling 2-1 that the Defense Department lawfully designated the AI startup a national-security supply-chain risk over its refusal to let Claude be used for autonomous weapons or mass surveillance.

A federal appeals court on Friday upheld the Pentagon's blacklisting of Anthropic from military contracts, with the 2-1 decision by the U.S. Court of Appeals for the District of Columbia Circuit coming in the lawsuit Anthropic filed challenging its March designation as a national-security supply-chain risk. Writing for the majority, Judge Gregory Katsas — joined by Judge Neomi Rao, both first-term Trump appointees — held the designation lawful because it rested on Anthropic's refusal to assent to a contract term, not on retaliation for the company's views. Judge Karen LeCraft Henderson, a George H.W. Bush appointee, dissented.

Judge Katsas wrote that the blacklisting had been lawfully "based on the company's refusal to assent to a contract term that the Department deemed essential," and the court rejected Anthropic's argument that the Pentagon had retaliated against it for its views on AI safety and ethics, its First Amendment-protected free-speech claim failing alongside it. Reuters, whose wire described the ruling as a victory for President Trump and Defense Secretary Pete Hegseth in their battle with the AI startup over safety guardrails, also reported the decision came after months of failed negotiations and public recriminations in which Hegseth and Trump accused the company of endangering American lives through ideological posturing.

The majority's reasoning

The court found it reasonable for the Pentagon to act after Anthropic refused to allow its products to be used for autonomous weapons or mass surveillance. "The Secretary raises the deeply sobering prospect of overly constrained AI models shutting down unexpectedly and thus causing important military operations to fail," Judge Katsas wrote, in language quoted by Reuters.

The Associated Press quoted a second passage from the majority opinion: "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." The opinion noted that, by Anthropic's own admission, the company encodes restrictions into Claude that prevent the model from performing tasks Anthropic wishes to prevent.

The majority added that it had "no reason to doubt" Anthropic acts with "noble intentions" in restricting Claude, citing its privacy and AI-safety concerns — but held that the definition of a supply-chain risk hinges "on what Anthropic does, not why Anthropic does it." The judges also said they had "no quarrel" with the conclusions of a separate San Francisco case, while finding that a "bad motive" on Anthropic's part is not required to support a supply-chain-risk designation.

What the ruling permits — and what it doesn't

The decision lets the Pentagon keep removing Claude models from its systems and bar the use of Anthropic products for Defense Department work, including prohibiting contractors from using them in department contracts. The department had begun removing Claude from its workflows earlier this year, and the ruling upholds its ability to do so. Anthropic's products are not barred from the federal government as a whole under this ruling — only from Defense Department work, a boundary that distinguishes it from the broader government-wide ban a California judge already blocked.

Anthropic said in a Friday statement that it respectfully disagrees with the decision but remains confident in its position and is considering its options, including review of the three-judge panel's ruling by the full appeals court. The startup told the court the designation has cost it billions of dollars in lost business and damaged its reputation ahead of a highly anticipated initial public offering — the company's asserted harm, not an independently established figure. An Anthropic spokesperson separately told Gizmodo that another federal court "has already held the government's parallel designation unlawful" and that the company is "considering all options, including further review." Its substantive defense, as reported by Reuters, is that AI is not yet reliable enough to be safely used in autonomous weapons, and that the company opposes domestic surveillance as a violation of fundamental rights.

The California split

Last month, a federal judge in San Francisco struck down a parallel Pentagon designation of Anthropic under a different law, finding the administration had unlawfully retaliated against the company for its views on AI safety. U.S. District Judge Rita Lin also blocked the government-wide ban on Anthropic and Hegseth's order barring military contractors from doing any business with the startup.

The two cases involve two distinct statutory designations, and each lawsuit challenges only one of them: the San Francisco case struck down the first, while Friday's D.C. Circuit ruling upheld the second. The split means the Pentagon's Defense Department-wide designation stands even as the broader government-wide measures remain blocked.

How the fight got here

The Pentagon designated Anthropic as a risk under two different laws in March 2026, after the company refused to remove guardrails against its technology being used for autonomous weapons or mass domestic surveillance. The unprecedented move followed months of negotiations that ended in failure.

The relationship had begun cooperatively: in July 2025 the department and Anthropic teamed up in a $200 million deal through which the startup agreed to prototype frontier AI capabilities advancing U.S. national security. The rupture came in February, when Anthropic CEO Dario Amodei said the company would not allow use of its models for domestic surveillance or fully autonomous weapons systems. Hegseth responded with an X post calling the stance "a master class in arrogance and betrayal," Trump directed every federal agency on Truth Social to immediately cease all use of Anthropic technology, and Anthropic filed two lawsuits against the department the following month.

What comes next

Anthropic could request a rehearing by the same panel, seek review by the full D.C. Circuit, or take the case to the Supreme Court. The company has said only that it is weighing its options; no filing has been reported. The Pentagon's top spokesman, Sean Parnell, wrote in a social-media post that Friday's ruling "completely validates the Department's position," according to the Associated Press; the White House and the Defense Department did not immediately respond to Reuters' emailed requests for comment.