WASHINGTON: A US federal appeals court upheld the Pentagon’s ban on Anthropic’s Claude AI models on Friday, keeping the company’s models out of military systems. D.C. Circuit Court
The US Court of Appeals for the District of Columbia Circuit found sufficient support for the Pentagon’s decision. Officials argued that using Claude in military systems could pose a national security risk. One judge dissented.
The dispute began after Anthropic refused to remove some restrictions on Claude’s use. Its safeguards covered lethal autonomous warfare and mass surveillance of Americans. By contrast, Pentagon officials wanted permission to use the technology for “all lawful uses”.
Anthropic challenged the resulting supply-chain risk designation. The company argued that the government lacked a sound legal basis and had violated its constitutional rights. Pasted text
The company could now seek reconsideration by the full appeals court. Alternatively, it could ask the US Supreme Court to review the case.
Read: UN AI Safety Warnings Dominate Security Council Session
Separately, a federal judge in California struck down broader restrictions on Anthropic in August. That case involved different statutory authority, so the D.C. appeals court ruling did not overturn it. The circuit ruling did not overturn the California decision. Pasted text
Pentagon Undersecretary Emil Michael welcomed the ruling on X, saying private companies should not insert their views into the military chain of command.
Read: Stop the AI Race Protest Targets OpenAI, Anthropic in SF
Anthropic can request an en banc review by the full 11-judge panel of the D.C. Circuit Court of Appeals or appeal the case directly to the U.S. Supreme Court.
The company could now seek reconsideration by the full appeals court. Alternatively, it could ask the US Supreme Court to review the case.
Separately, a federal judge in California struck down broader restrictions on Anthropic in August. That case involved different statutory authority.
DC and California courts based their rulings on different statutory authorities. The DC court did not reverse the California court ruling.
Read: AI Intelligence Error Nearly Triggered US Chinese Ship Raid
The Pentagon celebrated the ruling, with Undersecretary Emil Michael writing on X that “The hammer of justice has smashed @AnthropicAI arguments… Warfighters will sleep better knowing that no private company will insert their opinions in the chain of command.”
