Nonprofits sue to freeze frontier AI models
On November 12, 2026, four nonprofits sued Anthropic, Google, Meta, OpenAI, and SpaceXSI in San Francisco to freeze more capable models. No order yet.

On November 12, 2026, four nonprofits sued Anthropic, Google, Meta, OpenAI, and SpaceXSI in San Francisco to freeze more capable models. No order yet.

On November 12, 2026, Irreplaceable Education Fund, Public Citizen, Heat Initiative, and The Alliance for Secure AI sued in the Superior Court of California, County of San Francisco. Public Citizen's announcement ↗ names the defendants as Anthropic, Google, Meta, OpenAI, and SpaceXSI. The picture above is ours.
Counsel posted the complaint as Irreplaceable et al. v. Anthropic et al. ↗. The caption lists Irreplaceable Education Fund; Public Citizen, Inc.; Heat Initiative, Inc.; and Future Tech Insights Inc., doing business as The Alliance for Secure AI. Defendants: Anthropic PBC; Google LLC; Meta Platforms, Inc.; OpenAI Group PBC; OpenAI Foundation; SpaceX Super Intelligence LLC; and SpaceX Super Intelligence Corp. The case-number line on that copy is blank. Exhibit AI ↗ calls it active, filed November 12, not a class action, and says that as of the complaint no further proceedings had taken place.
Public Citizen says former New Jersey Attorney General Matthew Platkin filed it. The claim is California's Unfair Competition Law, Business and Professions Code section 17200. Platkin's line: "AI companies are causing a public nuisance by knowingly developing and deploying AI systems that pose catastrophic risk." The announcement frames a freeze on systems the groups say can escape human control, including what the headline calls the proliferation of bioweapons and cyberattacks. That is the plaintiffs' theory. No court has adopted it.
In the complaint's defendant section, Anthropic sells Claude, Google sells Gemini, Meta sells models such as Llama and Muse, and OpenAI sells ChatGPT. Of the SpaceX entities, it says SpaceX "develops, markets, and sells the AI model Grok," and alleges on information and belief that "prior to July 2026, when xAI became SpaceX, the company's AI operations occurred in California."
The prayer on the posted complaint asks for:
An immediate injunction forbidding development or deployment of AI systems more capable than the defendants' most capable systems as of the date of the complaint.
An immediate injunction forbidding the use of AI to autonomously conduct or direct research and development of more capable systems.
Injunctions against the practices the complaint calls a public nuisance and an Unfair Competition Law violation.
Reasonable attorneys' fees under Code of Civil Procedure section 1021.5, and other relief the court finds proper.
Just Security ↗ (Matthew Nagle, November 13) describes that as an immediate injunction barring the five companies from building or deploying systems more capable than their current most advanced models, on one Unfair Competition Law claim argued as a public nuisance, plus a bar on AI autonomously directing AI research. The complaint's cutoff is "as of the date of this Complaint."
TokenPost ↗ (November 12) says the claims have not been adjudicated and "No freeze has been ordered."
Phil Aroneanu, founder and executive director of Irreplaceable Education Fund, said: "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."
No injunction is reported in the sources above. A chat in ChatGPT, Claude, Gemini, or Grok is not, on this record, closed by a San Francisco order. The prayer asks to stop systems more capable than those as of the filing date, and to stop AI from autonomously directing that research. It does not ask, in those paragraphs, to recall Claude, Gemini, ChatGPT, or Grok. Meta remains a defendant for Llama and Muse, as the complaint describes them.
CBS News ↗, updated November 13, reported: "A Meta spokesperson for Meta declined to comment. The other tech companies targeted in the suit did not respond to a request for comment." That is CBS's account of its own requests, not a ruling.
Product context on this site: ChatGPT vs Claude vs Gemini and the Grok 4 review. Anthropic's roadshow is a different filing, the IPO range. A different Anthropic case, about Max plan ads, already has a November 6 order: Claude Max usage claims.
"Important clarification: This lawsuit (filed Nov 12 in SF Superior Court by nonprofits including Public Citizen and Alliance for Secure AI) doesn’t seek to stop all AI development." — Aim Predict ↗ (@aim_predict). The rest of that post is his summary of the prayer, not the prayer.
"California, obviously..." — Space Investor ↗ (@SpaceInvestor_D), November 12, before naming the same plaintiffs, court, and companies.
"Two very different pressure points hit AI governance in the last 24 hours: one in a San Francisco courtroom, and one on the table at the EU's first dedicated AI summit in Dublin." — Michael A Scott ↗ (@MikeScott614), November 13. This story does not re-report the Dublin half.
Reddit was not reachable for this story.
The apps still open. Nothing in the complaint copy, Public Citizen's announcement, TokenPost, Just Security, Exhibit AI, or CBS's November 13 report is an order freezing a model.
The request is a cap tied to the filing date, not a recall of Claude, Gemini, ChatGPT, or Grok. Aim Predict's "doesn't seek to stop all AI development" matches that shape. "Current most advanced models" is Just Security's paraphrase. The complaint says "as of the date of this Complaint." A judge has not granted either wording.
Do not confuse this with the Max advertising suit. That case is about plan copy and already has an order. This one is a development freeze under the Unfair Competition Law. The claims are not adjudicated. As of CBS's November 13 story, Meta had declined to comment and the other named companies had not answered CBS.