An American judge has temporarily blocked the Pentagon’s effort to label Anthropic a “security problem” — and it’s a rather significant moment at the crossroads of AI, politics, and business. Judge Rita Lin paused the designation of Anthropic as a “supply‑chain risk” to national security, which would have effectively cut the company off from government contracts, and suggested that the Trump administration’s move may have been more of a punishment for the company’s public stance on AI safety than a genuine effort to protect military systems.

At the heart of the dispute is that Anthropic refuses to have its models run in autonomous weapons or for domestic surveillance — and the government responded by placing it on a list of risky suppliers without giving the company a real chance to defend itself. Anthropic framed the case around the Constitution: an infringement on free speech and the right to due process. The court is now saying: stop, at least until the entire proceeding is properly reviewed.