Artificial Intelligence
Appeals court upholds Pentagon ban on Anthropic
A Washington appeals court upheld the Pentagon’s ban on Anthropic in a 2-1 ruling, leaving the company to pursue rehearing options after a separate federal court struck down a related designation.
- A panel of the U.S. Court of Appeals for the District of Columbia upheld the Pentagon’s blacklisting of Anthropic in a 2-1 decision on Friday.
- Judge Gregory Katsas, joined by Judge Neomi Rao, said the continued use of Claude in Defence Department systems posed a statutorily covered national-security risk, while Judge Karen LeCraft Henderson dissented.
- The Department of Defense labelled Anthropic a supply chain risk in March, which blocked the military and its contractors from using Claude for work with the agency.
- Anthropic sued the Trump administration in both San Francisco and the D.C. Circuit in March, and a federal judge in San Francisco later ruled one designation unlawful.
- The appeals court delayed the ruling taking effect so Anthropic could seek rehearing or ask the Supreme Court to review the case.
A federal appeals court in Washington upheld the Defence Department’s ban on Anthropic on Friday. The 2-1 ruling rejected the company’s claim that the action was unauthorised, arbitrary and unconstitutional.
Judge Gregory Katsas, joined by Judge Neomi Rao, said the department had enough support to conclude that Claude’s continued use in its systems created a national-security risk covered by statute. Judge Karen LeCraft Henderson dissented.
The department had labelled Anthropic a supply-chain risk in March, which stopped the military from using the company’s models and blocked defence contractors from using them for work with the department. Anthropic had also signed a $200 million contract with the Pentagon in July 2025, but later talks over deployment on the GenAI.mil platform broke down.
Anthropic said it disagreed with the decision and was considering further steps, including rehearing and Supreme Court review. The panel delayed the ruling taking effect while those options remained open.
Named in this story
People
- Judge Gregory Katsas
- wrote the majority opinion
- Judge Neomi Rao
- joined the majority opinion
- Judge Karen LeCraft Henderson
- dissented
- President Donald Trump
- was part of the administration Anthropic sued and criticised Dario Amodei
- Pete Hegseth
- the defence secretary who accused Anthropic of trying to seize veto power over military decisions
- Dario Amodei
- Anthropic’s chief executive whom Trump criticised
Companies
- Anthropic
- challenged the Pentagon’s blacklisting of its models
Governments and agencies
- Department of Defense
- labelled Anthropic a supply chain risk and barred use of Claude
- U.S. Court of Appeals for the District of Columbia
- upheld the blacklisting in a 2-1 ruling
- U.S. District Court in San Francisco
- heard Anthropic’s challenge and later ruled one designation unlawful
- Trump administration
- was sued by Anthropic over the blacklisting
- Supreme Court
- could be asked to review the case
Products and systems
- Claude models
- the AI models at the centre of the dispute
- GenAI.mil AI platform
- the Pentagon platform on which deployment talks broke down
How the source tells it
The piece is mostly a straight legal report, but it uses conflict-heavy and national-security framing to raise the stakes.
- conflict loaded setback language and repeated emphasis on the dispute with the Trump administration
- alarm national-security wording and the judge’s focus on risk present the issue as dangerous