Appeals Court Upholds Pentagon's Anthropic Risk Label — SkimNews
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- DC federal appeals court ruled 2-1 on Friday to uphold one of the Pentagon's supply-chain risk designations against Anthropic, finding the department had "ample support" that integrating Claude into DoD information systems posed a national-security risk.
- The majority opinion leaned on Anthropic's own admission that the company "encodes restrictions into Claude that prevent the model from performing tasks that Anthropic wishes to prevent" — the same refusal to allow autonomous weapons or domestic surveillance that Secretary of Defense Pete Hegseth deemed a significant national security risk.
- The ruling leaves one designation in place indefinitely; a separate San Francisco federal judge already tossed the other supply-chain risk label in March and confirmed that decision last month, with both cases facing years of appeals before final resolution.
- Anthropic spokesperson Danielle Cohen said the company remains confident and is weighing an appeal to the full DC Circuit or the US Supreme Court; the panel also rejected Anthropic's due process and free speech claims, calling the dispute "standard contract negotiations."
- The Pentagon is replacing Claude with alternatives including SpaceX's Grok, Google's Gemini, and OpenAI's GPT models, though the article notes some Google and OpenAI employees have objected to their employers striking military deals Anthropic rejected on ethical grounds.
- Anthropic reported losing revenue immediately after the designations as customers worried about doing business with a "government pariah," but the company has since touted growing sales and is moving toward a potential IPO later this year.
Why it matters: The 2-1 ruling keeps Anthropic effectively locked out of Pentagon contracts on one of two legal tracks at a pivotal moment — the company is preparing for a potential IPO later this year, and the court's framing of Anthropic's safety guardrails as evidence of a national-security risk sets a precedent that could chill other AI vendors from encoding similar restrictions on military clients.
Ask SkimNews
