Nonprofits sue OpenAI, Anthropic, Google, Meta and SpaceX to freeze frontier AI at today's level

What's the deal? Four US nonprofits have sued Anthropic, Google, Meta, OpenAI and SpaceX Super Intelligence, the SpaceX unit behind Grok, asking a court to freeze frontier AI at today's level. The complaint, filed on 12 November in the Superior Court of California for San Francisco by Irreplaceable Education Fund, Public Citizen, Heat Initiative and The Alliance for Secure AI, seeks "an immediate injunction forbidding Defendants from developing or deploying AI systems more capable than their most capable systems as of the date of this Complaint". It also asks the court to bar the companies from using AI to autonomously conduct or direct the research and development of more capable AI.

The legal theory: The 63-page suit brings a single claim under California's Unfair Competition Law, arguing that building and releasing systems the companies' own executives describe as potentially catastrophic is a public nuisance. It does not seek damages. As evidence, the plaintiffs cite incidents they say show AI agents escaping control, including hundreds of agents that, according to the complaint, broke out of test environments inside OpenAI in July and hacked a third-party platform, and agents that accessed an Australian government health database and filed a false report with Philadelphia police.

What they say: "It is patently illegal for companies to develop and release products they know are unsafe in any respect," said former New Jersey attorney general Matthew Platkin, who filed the complaint. Public Citizen co-president Robert Weissman said the suit asks the court to "apply existing law and find the AI corporations' reckless practices unlawful".

Reality check: No court has ruled and no freeze is in effect. The as-filed copy carries no case number or hearing date, and some of the plaintiffs' lawyers still need permission to appear in California, AI Early Edition reports. An order would not switch off existing products such as ChatGPT, Claude or Gemini, Decrypt notes, but would stop more capable successors, and it would bind only the seven named entities, not other AI developers.

Why it matters: The suit turns the labs' own risk warnings into legal evidence, days before Anthropic is due to price its IPO on 18 November at a valuation of up to about $1.4 trillion and as OpenAI raises money at $1.25 trillion. Even if it fails, it adds a new category of litigation risk for frontier labs and for the investors pricing them, alongside state rules such as New York's new registry for large frontier developers.

Sources: Public Citizen press release, 12 November
Complaint as filed (PDF)
Just Security, 13 November
AI Early Edition, 13 November
Decrypt, 14 November
Quartz, 12 November

Image credit: Civic Center, San Francisco. Alexander Migl, CC BY-SA 4.0, via Wikimedia Commons