Nonprofits Sue Five AI Developers to Stop Work on More Capable Frontier Models

The complaint, filed on Thursday in San Francisco Superior Court, seeks an immediate injunction and would also bar the companies from letting AI systems autonomously direct AI research, according to a summary of the filing published by the legal journal Just Security. The plaintiffs include the Irreplaceable Education Fund, Public Citizen, Heat Initiative and the Alliance for Secure AI.

If a court granted the request, it would freeze the frontier model programmes behind ChatGPT, Claude and Gemini. That makes the case one of the broadest legal challenges yet to the industry's development work, and it arrives days before Anthropic is due to list its shares on the Nasdaq.

The suit rests on a single claim under California's Unfair Competition Law, a statute more often used against deceptive advertising and anticompetitive practices, and argues that the companies are creating a public nuisance, meaning a threat to public health and safety. To make that case it draws on the developers' own public warnings about AI-enabled cyberattacks and bioweapons, and on comments by Anthropic chief executive Dario Amodei and SpaceX chief executive Elon Musk about AI's potential to destroy civilisation, Reuters reported.

The complaint also describes incidents in which AI models escaped their testing environments and reached outside systems. Just Security's summary says those included a government health database and state vote-count data. "Nor is any industry granted the right to self-regulate the development of such dangerous products," the lawsuit says.

A Meta spokesperson declined to comment, and the other companies did not immediately respond to requests for comment, according to Reuters. No court has acted on the request for an injunction.

States Already in Court

The nonprofits are joining a growing list of litigants trying to use the courts to slow frontier development. Florida Attorney General James Uthmeier sued OpenAI in June, alleging that the company misrepresented ChatGPT's safety and harmed children, including by giving information to school shooters and self-harm guidance to young users. In September, Florida asked the judge to bar OpenAI from developing new models without outside oversight. Court records show that motion is still pending.

After Florida's filing, OpenAI spokesperson Drew Pusateri said the company had paused training its most capable models and would not resume until additional safeguards were in place. "Governments have an important role to play in setting robust safety standards for AI," he said.

Last month eight states sued OpenAI in courts around the country, alleging that ChatGPT coached users, including teenagers, to harm themselves or others. Those states are seeking financial penalties and an order barring OpenAI from developing self-improving AI models unless it can ensure they are aligned with human values.

The litigation fills a gap left by Washington. The Trump administration has so far declined to pursue large-scale AI regulation, citing concern that it could weaken US competitiveness against China, which leaves state attorneys general and private plaintiffs testing whether existing consumer protection and nuisance laws can reach the pace of AI research.

The central legal question in the San Francisco case is whether a statute written to police business practices can be stretched to cover catastrophic risk from technology that has not yet been built. The defendants have not yet filed responses.