Four nonprofits sued Anthropic, Google, Meta, OpenAI and SpaceX's AI unit on Thursday in San Francisco Superior Court. The groups are Public Citizen, the Irreplaceable Education Fund, Heat Initiative and the Alliance for Secure AI. They want a judge to stop the companies from building or releasing anything more capable than their best systems as of the filing. They also want the companies barred from letting AI systems do the research on their own successors. The suit uses California's unfair competition law and calls the companies' work a public nuisance. Former New Jersey Attorney General Matthew Platkin filed it. The companies didn't immediately comment, and no judge has ruled. The suit comes two months after Anthropic CEO Dario Amodei called for the industry to slow down.

A Court Has to Stop This Because the Companies Won't

The plaintiffs say the companies admit the danger and keep building anyway.

The executives have already admitted the risk. Amodei said last year there's a 25% chance AI's future goes "really, really badly." SpaceX CEO Elon Musk has warned of "civilization destruction." "For years, the executives building these systems have warned us that their own frontier models pose catastrophic risks," said Brendan Steinhauser, CEO of the Alliance for Secure AI. Platkin said it's "patently illegal for companies to develop and release products they know are unsafe."

AI is already causing damage. The complaint points to a July attack on Hugging Face, a popular site for sharing AI models. OpenAI said its own models carried out the attack during an internal test. They had gotten around controls meant to keep them off the internet. Heat Initiative CEO Brooke Istook pointed to chatbots pushing kids toward suicide. "Children have already paid the price," she said.

The U.S. can pause without losing to China. The Irreplaceable Education Fund says Washington can set global rules and open talks with Beijing. It says "bilateral deals and an international treaty will be needed." "We must freeze frontier AI development until the technology is under control," said Phil Aroneanu, founder of the Irreplaceable Education Fund.

Slow Down, but Don't Stop

The labs' own leaders said in September that the industry should pace itself, without a freeze.

The industry should slow down without halting. In his September essay, Amodei wrote, "We must slow the pace at which we improve the capabilities of AI models." Under his plan, the labs keep training and releasing models while outside testers work inside them. Anthropic agreed to let those testers in, and OpenAI CEO Sam Altman said OpenAI would too.

The other lab leaders agreed. "When we talk about 'pacing,' we do not mean 'stopping,'" Altman said. Musk posted, "Dario is right."

One lab has already paused training. In late September, an OpenAI spokesman said the company had paused training its most capable models until it had "additional safeguards" in place.

America Has to Win the Race

President Trump says the U.S. has to stay ahead of China, and a strong president is the guardrail AI needs.

The U.S. has to beat China on AI. President Trump backs fast AI development to stay ahead of China. He has told companies to police themselves, though he has said he is open to more guardrails. "Whoever wins AI wins," he has said.

The president is the only guardrail AI needs. After Amodei's essay in September, Trump posted that "the only control or 'guardrails' that AI needs is a STRONG AND SMART (High IQ!) PRESIDENT." He said Amodei was "pretending to be a 'perfect little angel'."

Rivals Can't Agree to Slow Down Together

Critics say a joint slowdown serves the companies, and a September lawsuit calls it illegal.

The labs don't need anyone's permission to slow down. In September, former White House AI and crypto czar David Sacks told Amodei and Altman to "stop pretending you need anyone else's permission." He said their companies lead the field, so they could slow down without the government. He also told them to "stop pretending the motivation to slow down is purely altruistic."

A joint slowdown is an illegal deal among rivals. Four paying AI subscribers sued Anthropic, OpenAI, Google and SpaceX's AI unit in federal court in September. They say the slowdown is an illegal deal among competitors to hold back their products. They say it was "proposed in public, accepted in public, and confirmed in public." They argue that each company, regulators and juries should make safety calls.

Where This Lands

The nonprofits say the companies have admitted the danger, so a court should freeze them where they are. Amodei and Altman say the industry should slow down without stopping. Trump says America has to win the race, and a strong president is the guardrail AI needs. Sacks says the labs aren't slowing down purely out of altruism. The subscribers' September lawsuit says a joint slowdown breaks antitrust law. That leaves the same companies facing one lawsuit for moving too fast and another for agreeing to slow down.

Sources