Lawsuits to slow frontier AI: a dated tracker of the nonprofit, antitrust and state cases

Courts are becoming a second front in the argument over how fast frontier AI should move. This tracker logs every case we can verify that asks a judge to slow, pause or restrict frontier AI development. It is a living page: we will update it through corrections as filings and rulings arrive. Status is as of November 14, 2026.

The four cases at a glance

The cases make different, even opposing, arguments. Three ask a court to restrict development on safety or consumer harm grounds. One argues the labs illegally agreed among themselves to slow down.

• Nonprofit injunction suit, filed November 12, 2026, San Francisco Superior Court.
• Consumer antitrust class action, filed September 18, 2026, Northern District of California.
• Florida attorney general's suit against OpenAI, filed June 2026, with a September 28 motion pending.
• Eight state suit against OpenAI, filed October 2026.

Case 1: the nonprofit injunction suit

Plaintiffs: the Irreplaceable Education Fund, Public Citizen, Heat Initiative and The Alliance for Secure AI, with former New Jersey Attorney General Matthew Platkin filing for them, per Tech Times. Defendants: OpenAI, Anthropic, Google, Meta and SpaceX's AI unit.

Claims: California's Unfair Competition Law and public nuisance, among other theories, per Dataconomy and Tech Times. Relief sought: an immediate injunction barring the defendants from building or deploying AI more capable than their current most advanced models, and from letting AI autonomously direct AI research, per Just Security. Dataconomy quotes the complaint: "No other industry is allowed to treat its own development of unsafe products."

Status: no defendant had responded and no ruling or hearing date had been reported as of November 12. Meta declined to comment. Our full report is in the news piece published alongside this tracker.

Case 2: the antitrust class action over a pace agreement

Filed September 18, 2026 in the U.S. District Court for the Northern District of California, San Francisco Division, per Unite.AI. Plaintiffs Charles Buist, Nick Spetsas, Cheyenne Hunt and Christine Bullock sue for themselves and a proposed nationwide class. Defendants: Anthropic PBC, OpenAI OpCo LLC, SpaceXAI LLC and Google LLC. Counsel includes Nicholas C. Rowley of Trial Lawyers for Justice.

Claim: the defendants allegedly agreed to slow the improvement of their competing frontier AI products, which the complaint treats as an output restriction in violation of Section 1 of the Sherman Act. The complaint argues the restraint is unlawful per se and, alternatively, under quick look and rule of reason analysis. It alleges coordination began before a September 12, 2026 public exchange, including a working group formed in July 2026, and defines the market as U.S. paid consumer subscriptions to general purpose frontier AI assistants, where the four defendants allegedly hold at least 80%. Per Unite.AI, the plaintiffs say they do not challenge independent safety measures, unilateral decisions about pace, or lobbying. The coverage quotes an Amodei essay: "We must slow the pace at which we improve the capabilities of AI models."

Relief sought: treble damages under the Clayton Act, an injunction against horizontal agreements on pace, compute or training limits, coordinated release delays and capability checkpoints, and a jury trial.

Case 3: Florida v. OpenAI

Florida Attorney General James Uthmeier sued OpenAI in June 2026, alleging ChatGPT misrepresented its safety and harmed children, per Dataconomy. A September 28 motion in Highlands County asks the court to bar OpenAI from developing new models without outside oversight, per Tech Times. Status: pending.

OpenAI spokesperson Drew Pusateri said the company had paused training its most capable models until additional safeguards are in place, as paraphrased by Dataconomy and Tech Times. It is a paraphrase, not a verbatim quote, and we have not verified the pause independently.

Case 4: the eight state suit

Eight states sued OpenAI last month over alleged ChatGPT coaching of users, including teens, toward self harm or harm to others, per Dataconomy. They seek financial penalties and an order barring development of self improving models unless they can be shown to be aligned with human values. Status: no ruling reported in the sources we reviewed. The chatbot cases are tracked in more detail in our state attorney general chatbot lawsuit tracker.

What these cases have in common, and where they differ

All four try to use existing law because no federal statute regulates frontier development directly. They differ on the remedy: a freeze, a ban on self improving models, an end to coordination, or damages. The Buist case is the odd one out, since it treats slowing down as the offense. If both it and the nonprofit suit advanced, the labs could face one court saying coordination to slow is illegal and another asking them to stop altogether.

Physical infrastructure has its own docket. See our data center backlash tracker for moratoria and local suits.

What to watch next

• Whether a judge sets a hearing on the nonprofit suit's emergency injunction request.
• Motions to dismiss in the Buist case and the identity of the judge.
• A ruling on Florida's September 28 motion.
• Any lab answering on the merits rather than moving to dismiss.

How we will keep this current

When a case changes status we will re-send this page with a dated correction, not publish a new article. If you see a filing we missed, we want it. We only log cases we can source to a docket, a complaint or two independent reports.

Our take

None of these cases has produced a slowdown, and most legal experts would be right to call the injunction request a long shot. The reason to track them anyway is that they set the vocabulary for the next round of regulation: nuisance, unfair competition and coordination. Even losing suits create discovery and public records about what the labs knew and when. That, more than any injunction, is what to watch.