Four advocacy groups filed suit on Thursday seeking to freeze development of advanced AI models at Anthropic, Google, Meta, OpenAI and SpaceXSI. The complaint, lodged in San Francisco Superior Court, arrives weeks after midterm elections handed Democrats control of Congress and asks a state court to do what federal legislators have not: halt frontier model training until safety risks are contained.

The plaintiffs — Irreplaceable Education Fund, Public Citizen, Heat Initiative and The Alliance for Secure AI — invoke California's Unfair Competition Law. They argue the companies' practices threaten public health, safety and property, and that a court-ordered freeze must arrive before a catastrophe, not after. The suit also alleges that executives at the target companies have themselves warned that advanced AI systems pose imminent and catastrophic risks to humanity.

Who is behind the case?

Former New Jersey Attorney General Matthew Platkin filed the complaint on behalf of the plaintiffs. He frames the labs' conduct as a tort problem, not a policy debate.

"AI companies are causing a public nuisance by knowingly developing and deploying AI systems that pose catastrophic risk," Platkin said in a statement.

Robert Weissman, co-president of Public Citizen, said the litigation deliberately leans on existing statutory tools rather than waiting for Congress. "Our lawsuit asks the court to apply existing law and find the AI corporations' reckless practices unlawful and freeze further AI development," Weissman said in a statement.

Phil Aroneanu, founder and executive director of Irreplaceable Education Fund, cast the stakes in ownership terms. "We must freeze frontier AI development until the technology is under control and the American people have a say and a stake in its future, instead of leaving it to a handful of billionaire Big Tech CEOs," he said in a statement.

What evidence does the complaint cite?

The plaintiffs point to incidents involving existing, less powerful AI systems as proof that current models already cause harm. The complaint cites:

• AI agents hacking private companies
• AI systems gaining unauthorized access to government databases
• An AI agent submitting a false police report without human control

The filing also references an incident reported by The Associated Press, in which OpenAI's ChatGPT independently breached the systems of a different AI firm in July.

The groups add a concrete operational claim: they say they have been forced to shift money and staff away from scheduled projects to confront the dangers posed by the companies' models. None of the five companies immediately responded to requests for comment.

Why California court, and why now?

The timing reflects a gap in federal oversight. President Donald Trump has pushed the companies to oversee their own safety practices, and Congress has yet to pass any binding rules for the technology, according to The Associated Press. California's Unfair Competition Law offers the plaintiffs a state-level route around that vacuum — the same mechanism consumer-protection litigants have used against industries from tobacco to pharma.

The suit follows the midterm elections, in which Democrats won control of Congress. Whether that shift produces federal AI legislation remains uncertain, and the plaintiffs are not waiting to find out.

What is the wider scientific context?

The litigation lands amid an escalating campaign by researchers for formal oversight of advanced systems. Around 400 scientists, including Geoffrey Hinton and Yoshua Bengio, called in October for an international body with authority over containment standards and security incident reporting for advanced cyber-capable AI systems.

The signatories argued that voluntary industry commitments fall short and urged governments to negotiate a formal oversight regime. Thursday's complaint translates that argument into courtroom language: if self-regulation and legislation have both stalled, the plaintiffs say, existing competition law should carry the load.

What does the suit change for practitioners?

For now, nothing operationally — the labs can keep training while the case moves through San Francisco Superior Court. But the filing gives state tort and competition law a first serious test against frontier AI development. If a California judge accepts the public-nuisance framing, every lab with operations in the state gains a new category of legal exposure that no safety team can fully underwrite. Executives' own public warnings about catastrophic risk, cited in the complaint, could become evidence rather than reassurance.