undefined | Anthropic loses appeals court bid to temporarily block Pentagon blacklisting

A federal appeals court in Washington, D.C., rejected Anthropic’s request for an emergency stay that would have halted the Department of Defense’s decision to label the AI company a “supply‑chain risk” and effectively blacklist its technology. The court said the balance of equities favored the government, noting that the alleged harm to Anthropic was largely financial, whereas the DOD’s action involved protecting vital AI capabilities during an active military conflict. Consequently, the court denied the motion for a stay while the case proceeds on its merits.

The Pentagon’s designation, issued in early March, obliges defense contractors to certify that they do not use Anthropic’s Claude models in any work for the military. Anthropic argued that the label was an unconstitutional, arbitrary retaliation that threatened its free‑speech rights and could cause irreparable damage. While the judges acknowledged the company could suffer some immediate harm, they concluded that its interests were primarily financial and that “substantial expedition is warranted” for the government’s national‑security concerns.

The legal battle follows a separate San Francisco federal court ruling that granted Anthropic a preliminary injunction blocking the Trump administration’s ban on using the Claude AI system. The company’s efforts to secure a broader injunction against the DOD’s blacklist have thus far been unsuccessful, and the dispute now moves forward in the appellate system. The case highlights the growing tension between emerging AI firms and U.S. defense agencies over control, access, and the ethical use of powerful generative‑AI technologies.

Read more: undefined

#anthropic #pentagon #departmentofdefense #appealscourt #federalcourt