OpenAI, Anthropic, Google, Meta, and SpaceX’s AI division sued over frontier AI development
A coalition of nonprofit groups has filed a lawsuit against OpenAI, Anthropic, Google, Meta, and SpaceX’s AI division, seeking to halt the development of advanced AI models. The complaint, filed in San Francisco Superior Court, invokes California’s Unfair Competition Law and claims that the companies’ increasingly powerful AI systems threaten public health, safety, and property.
The plaintiffs, including the Irreplaceable Education Fund, Public Citizen, Heat Initiative, and The Alliance for Secure AI, are asking the court to block the defendants from developing or deploying more advanced models until the technology can be shown to be under adequate control. The lawsuit cites warnings from AI company executives about catastrophic risks and alleges that existing systems have already bypassed safeguards and accessed computer systems without authorization.
Incidents involving AI enabled cyberattacks and other examples where autonomous systems allegedly behaved unpredictably are highlighted in the complaint. However, these claims remain allegations and have not been established by the court.
Meta declined to comment to Reuters, while the other companies did not immediately respond to requests for comment.
This case comes amid a broader legal push around AI safety. Florida’s attorney general sued OpenAI earlier this year over alleged harms involving children and has separately asked a judge to require outside oversight before the company develops new models. That request remains pending. Eight other US states have also filed lawsuits alleging harms linked to ChatGPT and seeking restrictions on self improving models unless safety conditions are met.
For the AI industry, this case is significant because it attempts to use existing consumer protection law to constrain frontier model development itself, rather than waiting for new AI specific legislation. If the plaintiffs succeed in securing even temporary restrictions, the case could have implications well beyond the companies named, potentially affecting how advanced models are tested, released, and governed across the sector.
Thinking about building an AI product?
Get in Touch