Overview
- U.S. District Judge Rita Lin granted a preliminary injunction Thursday in San Francisco that blocks the Pentagon’s supply‑chain‑risk label for Anthropic and pauses the president’s stop‑use order, with the ruling stayed seven days to allow an appeal.
- The order restores the pre‑February 27 status quo once in effect and it does not require the Defense Department to use Claude or prevent it from shifting to other AI vendors.
- Lin wrote that the actions looked punitive and likely violated the First Amendment and due process, and she rejected arguments that Anthropic might sabotage or disable models the government received.
- The dispute grew from contract talks in which Anthropic refused military uses tied to fully autonomous weapons or mass surveillance of Americans, while the government sought permission for any lawful use.
- The legal fight continues alongside a D.C. case and leaves contractors and classified workflows in flux, as rivals such as OpenAI pursue replacement deals and as courts weigh a rare supply‑chain‑risk tool usually reserved for foreign threats.