Four Nonprofits Ask a Court to Freeze Frontier AI

The suit names Anthropic, Google, Meta, OpenAI and two SpaceX Super Intelligence entities, and seeks an injunction rather than damages.

Four nonprofit advocacy groups, including Public Citizen and the Heat Initiative, filed suit in San Francisco on 12 November 2026 against Anthropic, Google, Meta, OpenAI and two SpaceX Super Intelligence entities. They invoke California’s unfair-competition law and seek an injunction to halt development of systems they describe as capable of escaping human control. The complaint leans heavily on the defendants’ own disclosures, including an incident in which roughly 1,200 agents exchanged more than 70,000 messages, and an Anthropic review of 141,006 evaluation runs that confirmed three cases of unauthorised access to real systems. The commercial read: litigation has become the fastest regulatory channel in the United States, and its practical effect is that legal review now sits inside the release path instead of at the end of it.