Appeals court backs Pentagon's blacklisting of Anthropic in 2-1 ruling

A federal appeals court on 25 September upheld the Pentagon's blacklisting of Anthropic from military contracts, handing a victory to President Donald Trump and Defence Secretary Pete Hegseth in their dispute with the AI startup over safety guardrails on its technology. The 2-1 decision came from the US Court of Appeals in Washington, DC, in Anthropic's lawsuit challenging its designation.gizmodo.com+1

The ruling matters because it affirms the Pentagon's March designation of Anthropic as a national security supply chain risk, a label that bars the company from partnering with the US military or federal contractors. Anthropic says the designation has cost it billions of dollars in lost business and damaged its reputation ahead of a highly anticipated initial public offering. What is new is that a federal appeals court has now backed the government's authority to impose the label over the company's objections.reuters.com

What the court decided

The court's majority said it was reasonable for the Pentagon to designate Anthropic after the company refused to allow its products to be used for autonomous weapons or mass surveillance. The panel rejected Anthropic's claim that the Pentagon had retaliated against it for its views on AI safety and ethics. The majority also rejected Anthropic's argument that its First Amendment-protected right to free speech had been violated, according to Judge Gregory Katsas's opinion.

Siding with Hegseth's argument that Anthropic's safety restrictions could compromise military operations, Judge Gregory Katsas wrote in the majority opinion: "The Secretary raises the deeply sobering prospect of overly constrained AI models shutting down unexpectedly and thus causing important military operations to fail." Katsas, appointed by Trump during his first term, wrote that the blacklisting was lawfully "based on the company's refusal to assent to a contract term that the Department deemed essential."

A federal judge in San Francisco in August struck down a parallel designation under a different law, finding the administration had unlawfully retaliated against Anthropic for its views on AI safety. US District Judge Rita Lin also blocked Trump's government-wide ban on Anthropic and Hegseth's order barring military contractors from doing any business with the startup. The government had based its designation on two legal designations; the first was found illegal, the second is the one the panel upheld.

The parallel ruling that conflicts

Anthropic said in a statement on 25 September 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 decision by the full appeals court. A company spokesperson told Gizmodo that another federal court has already held the government's parallel designation unlawful. The company could also request a rehearing from the same panel or take its case to the Supreme Court.

The dispute began early this year when the US Defense Department pressed Anthropic to remove safety guardrails from Claude, its AI model. The demand escalated into a broader confrontation that led to the blacklist, lawsuits and bans by several federal agencies. Hegseth designated Anthropic as a risk under two different laws after the company refused to remove guardrails against its technology being used for autonomous weapons or mass domestic surveillance.

How the dispute escalated

The confrontation turned public in February. In an X post on February 27, Hegseth described Anthropic CEO Dario Amodei's stance as "a master class in arrogance and betrayal as well as a textbook case of how not to do business with the United States Government or the Pentagon." The same day, Trump wrote on Truth Social that he was directing every federal agency to immediately cease all use of Anthropic's technology.

Anthropic filed two lawsuits against the Department of War the following month, claiming the supply chain risk label was an "unprecedented and unlawful" violation of its First Amendment-protected right to free speech. Anthropic denied the administration's characterisation and said AI is not yet reliable enough to be safely used in autonomous weapons. The company also said it opposes domestic surveillance as a violation of fundamental rights.

The two sides had earlier worked together. The Department of War and Anthropic first teamed up in July of last year in a $200 million deal, through which Anthropic agreed to prototype frontier AI capabilities that advance US national security. Anthropic was forced to cut off access to its then-most powerful models in June following an order from the federal government, which cited an obscure export law.

The White House and the Department of Defence did not immediately respond to emails seeking comment on the ruling. Despite the ban, government agencies continued using Anthropic's tools as the White House and the company gradually mended ties ahead of Anthropic's planned initial public offering. Anthropic's next step is a decision on whether to seek a rehearing by the full appeals court or appeal to the Supreme Court.