TL;DR — Key Takeaways
- Four nonprofits sued OpenAI, Google, Anthropic, Meta and SpaceX’s AI division, SpaceXSI, in San Francisco Superior Court on Nov. 12. They’re seeking a court order to halt the development and deployment of more advanced AI models.
- The plaintiffs invoke California’s Unfair Competition Law and allege that developing and deploying advanced AI without adequate safeguards creates a public nuisance. They cite risks including cyberattacks, the spread of bioweapons and the loss of human control over AI systems.
- The case follows separate lawsuits against OpenAI by Florida and eight other states. Florida’s request for outside oversight of new model development was still pending as of Nov. 12.
A coalition of four nonprofit organizations has filed a lawsuit against five leading AI developers, seeking a court order that would prevent the companies from building or releasing new AI models over concerns that the technology poses major risks to public safety.
Filed in San Francisco Superior Court, the lawsuit names OpenAI, Google, Anthropic, Meta and SpaceX’s AI division, SpaceXSI. The plaintiffs argue that the companies are developing powerful AI systems without adequate safeguards, potentially exposing the public to harm ranging from cyberattacks to the loss of human control over autonomous technology.
The legal action is based on California’s unfair competition law, which prohibits certain business practices that harm the public. The plaintiffs also allege that the companies’ activities constitute a public nuisance.
The four organizations bringing the case are Public Citizen, the Irreplaceable Education Fund, the Alliance for Secure AI, and Heat Initiative.
The lawsuit’s central demand is remarkably all-encompassing. The plaintiffs demand a halt to the development and deployment of new advanced AI models by some of the industry’s largest companies. If granted, such an injunction could disrupt product roadmaps for ChatGPT, Google’s Gemini and Anthropic’s Claude.
The case arrives as AI developers are investing heavily in agentic systems capable of executing complex tasks with limited human supervision. Unlike conventional chatbots, these systems can interact with software and make operational decisions.
That autonomy is a compelling concern for the plaintiffs. The complaint describes incidents involving AI systems conducting unauthorized cyber activity, including attacks against government websites. It also cites a reported July incident in which ChatGPT independently compromised another AI company’s systems.
The plaintiffs argue that these incidents demonstrate weaknesses in existing safety controls. They also point to public statements from AI executives acknowledging the potential dangers of advanced models. Anthropic CEO Dario Amodei has estimated a 25% probability that AI development will have seriously harmful consequences, compared with a 75% probability of a positive outcome. Elon Musk, who leads SpaceX, warned in 2023 that AI could threaten civilization.
While no court has ordered a halt to AI development, and existing chatbot services remain available, the lawsuit does raise potential concerns for businesses using AI. Perhaps the most compelling question: will future model development be governed primarily by vendors’ internal safety policies or by legally enforceable requirements?
More AI Lawsuits Pending
The new lawsuit adds to a series of legal challenges seeking greater oversight of AI developers.
Florida Attorney General James Uthmeier sued OpenAI in June, alleging that ChatGPT exposed young users to harmful content and that the company misrepresented the platform’s safety. In September, Florida requested independent oversight before OpenAI could develop additional models. That request remains pending.
Eight other states filed separate lawsuits against OpenAI in October, alleging that ChatGPT encouraged harmful behavior among users, including teenagers. Those cases seek financial penalties and restrictions on the development of self-improving AI systems.
The federal government’s approach has been less restrictive. President Donald Trump’s administration has favored rapid AI innovation, partly to maintain U.S. competitiveness with China. Congress has also avoided comprehensive AI regulation. The Democrats’ victory in the November midterm elections could bring renewed efforts to regulate the technology, but the president has veto power and Democrats lack a veto-proof majority.
