A federal judge just ruled that the Trump administration broke the law when it blacklisted Anthropic's AI models over an alleged national security risk.
Judge Rita Lin's 59-page ruling found the government's "supply chain risk" designation against Anthropic was unlawful retaliation, not a legitimate security judgment. The label followed a lawsuit Anthropic filed in March, after a public fight with the Department of War and President Trump over the government's potential use of its models for surveillance and autonomous weapons. Trump had ordered an immediate blacklist of Anthropic products across federal offices, calling the company "leftwing nut jobs." Lin cited the government's later retreat from its own security claims, along with renewed talks between the two sides, as evidence the designation was punitive rather than genuine.
The ruling matters because "national security" has become a convenient label for punishing companies that pick fights with the administration, and courts have mostly stayed out of it until now. It lands just weeks after Anthropic was separately forced to pull its Claude Fable 5 model from some deployments under an export control directive - a reminder that a court win doesn't erase every administration action already underway.
Civil liberties groups are calling it a First Amendment marker. Anthropic and the federal government are still negotiating, so this fight probably isn't finished - it just lost its favorite excuse.