Skip to content
THURSDAY, NOVEMBER 19, 2026

Independently reported.

Tech

Nonprofits ask a judge to freeze frontier AI, while another suit calls a slowdown illegal

Four groups sued Anthropic, Google, Meta, OpenAI and SpaceX's AI unit on November 12. A September antitrust case argues the opposite: that the companies cannot agree to slow down.

By Theo Lindqvist and Rosa Delgado

· 3 min read

An empty courtroom with a judge's bench, a closed gavel and a tall stack of bound legal briefs in late afternoon light.
Illustration: Trestlewire

Two lawsuits against the same handful of AI companies now sit in San Francisco courts, and they ask for opposite things. On September 18, a group of plaintiffs including Buist sued Anthropic, OpenAI, Google and a SpaceX AI unit in federal court, alleging the companies illegally agreed to slow down. On November 12, four nonprofits sued Anthropic, Google, Meta, OpenAI and SpaceX's AI unit in state court, asking a judge to make them slow down.

The short version

Public Citizen, Heat Initiative, the Alliance for Secure AI and the Irreplaceable Education Fund filed suit in San Francisco Superior Court on November 12, seeking an injunction on more advanced AI models. No ruling, hearing date or company response has been reported. A separate September antitrust suit says a coordinated slowdown would itself be unlawful.

What the nonprofits filed

The complaint is titled "Complaint for Injunctive Relief for Violations of the Unfair Competition Laws," according to Public Citizen's release. It invokes California's Unfair Competition Law and alleges a public nuisance. Former New Jersey Attorney General Matthew Platkin filed it for the four groups. Platkin said it is "patently illegal" to develop and release products a company knows are unsafe.

The release says existing systems have already bypassed safeguards and reached outside computer systems without authorization. It lists AI agents hacking private companies, getting into government databases and submitting a false police report. Robert Weissman, Public Citizen's co-president, said the suit asks the court to "apply existing law."

The relief is broad. Al Jazeera reported that the groups want an emergency injunction on next-generation models, lifted only once the companies prove the technology is under human control. The Daily Caller, citing an Alliance for Secure AI press release, reported a second demand: a ban on recursive self-improvement, where models train themselves. Phil Aroneanu, who founded the Irreplaceable Education Fund, told Al Jazeera a real freeze has to specify what work stops and how compliance is verified.

The suit pointing the other way

Buist v. Anthropic was filed in the U.S. District Court for the Northern District of California, according to Straight Arrow News. It claims a Sherman Act violation: an agreement among competitors, the plaintiffs say, to reduce product quality and the rate at which products improve. Straight Arrow News reported that the suit followed an essay by Anthropic chief executive Dario Amodei, who wrote that the industry must slow the pace of capability gains.

The Buist plaintiffs argue that whether frontier AI should develop more slowly is a question for each company acting alone, or for Congress and the agencies. A pause the companies arrange among themselves is the thing they say is illegal.

The nonprofits ask for a pause that a judge imposes. The difference is who sets it. An order against named defendants is not an agreement among them, though neither report I reviewed says how a court would weigh the two suits against each other, and no judge has been asked to.

Why the timing matters

Al Jazeera ties the filing to the Democratic midterm win, which analysts say could tighten the rules around Silicon Valley. The plaintiffs urge the incoming Congress to start with a freeze and say courts must act because lawmakers have not. Against that, AI companies and President Trump have repeatedly framed the field as a race with China, and Trump has said "whoever wins AI wins."

Several things are not known. I could not read the 63-page complaint itself, so the descriptions above come from the plaintiffs' release and press coverage, which share a single origin. No case number or hearing date has been reported. Anthropic, Google, Meta, OpenAI and SpaceX did not immediately respond to requests for comment from the outlets that asked, and the coverage I reviewed offers no comparison with earlier cases that sought to stop a category of software development.

The first thing to watch is whether any defendant files a response in San Francisco Superior Court, and whether it argues the Buist suit makes a negotiated pause impossible.

  • Anthropic
  • OpenAI
  • Public Citizen
  • AI regulation
  • antitrust
  • Unfair Competition Law

About the reporter

Theo Lindqvist

Senior Correspondent, Technology & Business, Trestlewire

I got my first newsroom job in Seattle in 1999, which means I learned the trade covering companies that went public on a Tuesday and stopped existing by the following spring. That was a useful education. Since then I have covered the search wars, the rise of the app stores, the chip shortages, the antitrust cases against the biggest platforms, and the current argument about what artificial intelligence is actually worth.

Read full bio and all stories →

The Morning Wire

The day’s reporting, in your inbox before nine.

Free. Unsubscribe any time. We never share your address.