New York Times columnist Andrew Ross Sorkin and Anthropic CEO and co-founder Dario Amodei speak on stage at the 2025 New York Times Dealbook Summit at Jazz at Lincoln Center in New York on December 3, 2025.
Michael M. Santiago | Getty Images
A federal appeals court in Washington, D.C., on Wednesday denied Anthropic’s request to temporarily block the Pentagon’s blacklisting of artificial intelligence companies as a lawsuit challenging the sanctions unfolds.
The ruling came after a San Francisco federal judge late last month granted Anthropic a preliminary injunction barring the Trump administration from enforcing the Claude ban in a separate, related case.
“In our view, the fair balance here favors the government,” the appeals court said in its decision. “On the one hand, the risk of economic harm to a single private company is relatively contained. On the other is judicial control over how and through whom the Department of the Army secures critical AI technology during active military conflict. Therefore, we deny Anthropic’s motion for a stay pending review of the merits.”
In early March, the Department of Defense declared Anthropic an exposed supply chain risk, meaning the company’s use of its technology poses a threat to U.S. national security. The label requires defense contractors to certify that they are not using Anthropic’s Claude AI model in their work with the military.
Antropic asked the appeals court to review the Pentagon’s decision, arguing it was unconstitutional, arbitrary, capricious and a form of retaliation that did not follow procedures required by law, according to the filing.
In Wednesday’s ruling, the court acknowledged that Anthropic “is likely to suffer some irreparable harm if not suspended” but that the company’s profits “appear to be primarily financial in nature.” Although the company argued that the Pentagon was interfering with its right to free speech, “Antropic has not shown that its speech has become less lenient during the pendency of this litigation,” the order states.
Because of the damage Anthropic might suffer, the appeals court said, “a significant expedition is warranted.”
In a statement after the ruling, an Anthropic spokesperson said the company was “grateful that the court recognized the need to quickly resolve these issues” and was “confident that the court will ultimately agree that these supply chain designations are unlawful.”
Anthropic said, “While this lawsuit was necessary to protect Anthropic, our customers, and partners, we remain focused on working productively with the government to ensure all Americans have access to the benefits of safe and reliable AI.”
The Department of Defense relies on two different U.S. federal court designations to justify its supply chain risk litigation, which must be challenged in two separate courts.
Anthropic’s lawsuit against the Pentagon in March follows several dramatic weeks in Washington, D.C., between the Pentagon and one of the world’s most valuable private companies.
In a post on X in late February, Secretary of Defense Pete Hegseth declared Anthropic a supply chain risk, and the Pentagon soon notified the company of its formal decision in a letter. Anthropic is the first American company to be given this designation, which has historically been reserved for foreign adversaries.
Shortly before Hegseth’s post, President Donald Trump ordered federal agencies to “immediately cease” all use of Anthropic’s technology in a post on Truth Social. He said there will be a six-month phase-out period for government agencies like the Department of Defense.
The Trump administration’s actions surprised many officials in Washington, where Anthropic’s technology is embedded in many government agencies. The company was the first to deploy the model across the Department of Defense’s classified networks, championing its ability to integrate with existing defense contractors like Palantir.
Anthropic signed a $200 million contract with the Department of Defense in July, but negotiations stalled in September when the company began negotiations to introduce Claude to the Pentagon’s GenAI.mil AI platform.
The Pentagon wanted Anthropic to give the Pentagon 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 sides could not reach an agreement and the dispute went to court.
— CNBC’s Dan Mangan contributed to this report.
WATCH: Anthropic wins preliminary injunction in battle over Pentagon blacklist
