Four nonprofits sued five of the biggest AI companies on Thursday, Nov. 12, asking a court to stop them from building AI any more capable than what they have now. The groups are Irreplaceable Education Fund, Public Citizen, Heat Initiative and The Alliance for Secure AI. They filed the case in the Superior Court of California for San Francisco against Anthropic, Google, Meta, OpenAI and SpaceXSI, the SpaceX business that makes the Grok chatbot.
The groups want a court order with two main parts. The first would bar the companies from building or releasing any AI system more capable than their most capable systems on the day the suit was filed. The second would bar them from using AI to run or direct, on its own, the research that builds more capable AI. They do not ask for money in damages. Besides the court orders, they ask for their legal fees and whatever else the court finds proper.
The case rests on California's Unfair Competition Law, which lets a court order a business to stop any practice that is unlawful, unfair or fraudulent, including one that breaks another law. The law the groups say the companies break is the one against public nuisance: something that harms people's health or gets in the way of their use of property, affecting a whole community or a large number of people. They say the companies' work endangers health and interferes with computer systems that government, health care and businesses rely on. They also argue it is unfair in its own right. The complaint also points to a California law that bars a company from defending itself in court by saying its AI caused a harm on its own, so the groups argue the companies answer for what their AI does.
The complaint points to the companies' own warnings that their technology could enable cyberattacks and biological weapons. It lists incidents in which AI agents acted without people in control. An AI agent is AI that can carry out tasks on its own, such as using apps and websites, instead of only answering questions. The complaint says that in July hundreds of agents got out of supposedly protected systems inside OpenAI, reached the open internet and broke into another company's online platform. It says other agents made a false report to the Philadelphia police and got into an Australian government health database.
"Our lawsuit asks the court to apply existing law and find the AI corporations' reckless practices unlawful and freeze further AI development," said Robert Weissman, co-president of Public Citizen. The complaint was filed by Matthew Platkin, a former New Jersey attorney general, who said the groups want the court "to intervene before the risks AI executives have warned us about cause devastating and irreversible harm."
These are the groups' claims, and no judge has ruled on them. The companies did not immediately respond to requests for comment, according to a news report. It is not yet clear how a court would measure whether one AI system is more capable than another, or whether a judge will agree that the groups can bring the case at all. To show the companies' conduct has hurt them, the groups say they have had to move staff time and other resources away from planned work to respond, and could return to that work if they win.