Appeals Court Upholds Pentagon Blacklisting of Anthropic
A federal appeals court panel in Washington, D.C. upheld the Pentagon's blacklisting of Anthropic on Friday, dealing a blow to the AI company in its months-long battle with the Trump administration. The 2-1 decision rejected Anthropic's argument that the Department of Defense's ban on its Claude models was arbitrary, unauthorized, and unconstitutional[1].
The story reached 413 points on Hacker News with 733 comments, making it one of the most discussed stories of the day[2].
The ruling
Judge Gregory Katsas, writing for the majority, stated: "The Department had ample support for its conclusion that the continued integration of Claude into the Department's information systems, by the Department or its contractors, presented a statutorily covered national-security risk." Judge Neomi Rao joined the opinion. Both were appointed by President Trump. Judge Karen LeCraft Henderson, appointed by George H.W. Bush, dissented[1].
In March, the DOD labeled Anthropic a supply chain risk, a designation that purportedly threatens U.S. national security. This followed collapsed negotiations about how the military could use Claude AI models. The designation prevents the U.S. military from using Anthropic's models and blocks defense contractors from using them in their work with the agency[1].
How we got here
Anthropic signed a $200 million contract with the Pentagon in July 2025. When the company began negotiating Claude's deployment on the DOD's GenAI.mil AI platform in September, talks fell apart. The DOD wanted unfettered access to Claude across all lawful purposes. Anthropic wanted assurance that its technology would not be used for fully autonomous weapons or domestic mass surveillance[1].
They could not reach an agreement. Defense Secretary Pete Hegseth accused Anthropic of attempting to "seize veto power over the operational decisions of the United States military"[1].
The safety versus sovereignty debate
Katsas raised what he called the "deeply sobering" idea that "overly constrained" AI models could shut down unexpectedly, and the potential that Claude might be "subject to manipulation." While Anthropic refuted those claims, Katsas said decision-making authority ultimately rests with the President and the Secretary of Defense[1].
"In our Republic, it is the President and the Secretary of War who must determine how best to balance the competing risks," Katsas wrote. "In doing so here, the Secretary did not transgress any limits on his authority under the Supply Chain Security Act or the Constitution"[1].
The political dimension
Anthropic's relationship with the Trump administration has been fraught. Trump has repeatedly slammed CEO Dario Amodei on social media. Amodei drew Trump's ire by calling for an industry-wide slowdown, and was not invited to the state dinner the White House hosted for Chinese President Xi Jinping on Thursday[1].
"The Trump Administration has stopped AI 'people' from doing bad, or potentially bad, 'things,' like Dario (Anthropic!), who is now pretending to be a 'perfect little angel,' and we will continue to do so!" Trump wrote on Truth Social on Monday[1].
Two courts, two outcomes
Anthropic sued the Trump administration in two separate courts, challenging two distinct designations. A San Francisco federal judge ruled last month that one designation was illegal. The D.C. appeals court ruling on Friday upheld the second. Anthropic said it "respectfully disagrees" and is "considering all options, including further review"[1].
The appellate panel delayed the decision from taking immediate effect, giving Anthropic time to petition for a rehearing or seek an en banc rehearing by all judges on the D.C. Circuit Court of Appeals. Anthropic could also ask the Supreme Court to take the case[1].
Why it matters
This case sits at the intersection of AI safety, national security, and executive power. Anthropic refused to let its models be used for autonomous weapons or mass surveillance without guardrails. The Pentagon interpreted that refusal as a national security risk. A federal court has now agreed that the military can blacklist a company for insisting on usage restrictions[1].
The precedent is significant. If AI companies cannot set conditions on how their models are used by the military without risking blacklisting, the incentive structure for safety-conscious AI development shifts. Companies may either comply with any government demand to maintain contracts, or refuse to engage with the military entirely. Neither outcome serves the public interest particularly well[2].
Sources
[1] CNBC: "U.S. appeals court upholds Pentagon designation of Anthropic as supply chain risk" (September 25, 2026)
[2] Hacker News discussion (413 points, 733 comments)