Four nonprofits sue to halt the most advanced AI

Four nonprofit groups sued Anthropic, Google, Meta, OpenAI and a fifth company, "SpaceXSI," on Thursday, asking a judge to order a halt to the development of the most advanced artificial intelligence models.

The plaintiffs are Public Citizen, Irreplaceable Education Fund, Heat Initiative and The Alliance for Secure AI. They filed in California Superior Court in San Francisco County under the state's Unfair Competition Law. Their argument: models of this kind could slip out of human control and help others build biological weapons or launch cyberattacks, which puts public health, safety and property at risk.

The groups are seeking an injunction, a court order that would stop the companies before they go further. They are not asking to shut down products that already work. Phil Aroneanu of Irreplaceable Education Fund called for freezing frontier AI until the technology is "under control."

Matthew Platkin, a former New Jersey attorney general, is the plaintiffs' lawyer, and his firm, Platkin LLP, posted the complaint. Platkin called it "patently illegal" to develop and release products a company knows are unsafe. Robert Weissman of Public Citizen described the suit as a request that the court "apply existing law."

The plaintiffs' announcement does not say what SpaceXSI is or give a case number, and it carries no response from any of the five companies.

For users in Mexico, nothing changes for now: ChatGPT, Gemini, Claude and Meta's products keep running, and no order or hearing has been announced. If the judge granted the request, it would reach the new models that would power those services. The case is a state civil action, not federal regulation, and it hinges on whether the court accepts that an unfair competition statute covers how AI models are built.