A panel of the federal appeals court in Washington, D.C., on Friday upheld the Pentagon’s blacklisting of Anthropic, dealing a blow to the artificial intelligence company in its months-long battle with the Trump administration.
The 2-1 ruling rejected Antropic’s argument that the Pentagon’s ban on Claude models was arbitrary, unauthorized, and unconstitutional.
“The Department has sufficiently supported its conclusion that Claude’s continued integration into the Department’s information systems by the Department or its contractors poses a legally covered national security risk,” Judge Gregory Katsas wrote in a majority opinion for the U.S. Court of Appeals for the District of Columbia, which was joined by Judge Neomi Rao. Katsas and Rao were appointed by President Donald Trump.
Judge Karen LeCraft Henderson, appointed by former President George H.W. Bush, dissented.
In March, the Pentagon labeled Anthropic a supply chain risk, meaning the company posed a threat to U.S. national security after negotiations over how the military could use Claude’s AI models spiraled out of control. This designation prevents the U.S. military from using Anthropic models and also prevents defense contractors from using them in their work with Anthropic.
Since then, Anthropic’s relationship with the Trump administration has been troubled, with Trump repeatedly criticizing the company’s CEO Dario Amodei on social media. Mr. Amodei, who recently drew the ire of President Trump by complaining about an industry-wide slowdown, was not invited to a lavish state dinner hosted by the White House for Chinese President Xi Jinping on Thursday.
“The Trump administration has been stopping AI ‘people’ like Dario (human!), who is currently pretending to be a ‘perfect little angel’, from doing bad or potentially bad things, and we will continue to do so!” President Trump said in a post on Truth Social on Monday.
In March, Anthropic sued the Trump administration in the U.S. District Court of San Francisco and the U.S. Court of Appeals for the D.C. Circuit, seeking to cancel the blacklisting. Because the Department of Defense relied on two different designations to justify its supply chain risk actions, it had to litigate in two separate courts.
A federal judge in San Francisco ruled last month that certain designations were illegal. Friday’s ruling by the D.C. Court of Appeals upheld the second designation.
“We respectfully disagree with the court’s decision,” an Antropic spokesperson told CNBC in a statement. “Another federal court has already ruled that the government’s parallel designation is unlawful. We remain confident in our position and are considering all options, including further review.”
The appellate panel said it would delay the decision’s immediate effect to give Anthropic time to petition the same panel for reconsideration or seek an extensive rehearing of the case by all judges on the D.C. Circuit Court of Appeals. Antropic may also ask the Supreme Court to take the case.
Aerial view of the Pentagon, home to the U.S. Department of Defense headquarters, in Arlington, Virginia, on May 31, 2026.
Daniel Slim | AFP | Getty Images
Anthropic served as an early partner for many U.S. government agencies, including the Department of Defense, until the conflict between the two companies became public knowledge earlier this year. Anthropic was awarded a $200 million contract with the Department of Defense in July 2025, but negotiations broke down when the company began negotiations to introduce Claude to the Pentagon’s GenAI.mil AI platform in September of the same year.
The Pentagon wanted Anthropic to allow the military unfettered access to its models for all lawful purposes, but Anthropic wanted assurances that its technology would not be used for fully autonomous weapons or domestic mass surveillance. The two countries were unable to reach an agreement, with Defense Secretary Pete Hegseth accusing Antropic of attempting to “grasp a veto over U.S. military operational decisions.”
“The Department had reasonable concerns that Anthropic would manipulate Claude’s design to prevent it from performing national security functions that the Department was contractually authorized to do and determined necessary,” Katsas wrote Friday.
WATCH: CEOs of OpenAI, Anthropic and Hugging Face speak to UN Security Council
