The Pulse
Four AI Companies Face Suit Over Alleged Slowdown Pact
Four consumers sued Anthropic, OpenAI, SpaceXAI and Google in federal court, alleging the companies coordinated to slow AI development. The proposed class action claims the alleged agreement reduced the value of paid ChatGPT, Claude, Grok a
AI.info Team ·
Four consumers, four AI companies, one antitrust claim
Four consumers sued Anthropic, OpenAI, SpaceXAI and Google on September 18, accusing the companies of making an illegal agreement to slow the development of their competing artificial-intelligence products. The case was filed in the U.S. District Court for the Northern District of California and seeks to represent a nationwide class of paid subscribers.
The plaintiffs subscribe to ChatGPT, Claude, Grok or Gemini. Their complaint argues that coordinated restraint on model improvements would reduce what customers receive for their monthly payments, violating Section 1 of the Sherman Antitrust Act.
The companies have not been found liable, and the allegations have not been tested in court. Representatives for Anthropic, OpenAI, Google and SpaceXAI did not immediately respond to requests for comment, according to The Associated Press.
The alleged agreement took shape in public
The lawsuit centers on events from September 12, when Anthropic CEO Dario Amodei published an essay calling for industry cooperation to slow the pace of AI development while safety measures catch up. OpenAI CEO Sam Altman, SpaceXAI CEO Elon Musk and Google DeepMind co-founder Demis Hassabis each publicly responded to the proposal that day.
The complaint treats those public statements as evidence of coordination among competitors. It also points to a July 2026 statement signed by employees from several AI companies that acknowledged “intense competitive pressure not to unilaterally slow” development and called for government support for an international effort to pace automated AI progress.
Amodei’s essay proposed third-party evaluation of advanced models, common safety standards and limits on unchecked progress. The essay also acknowledged that cross-company coordination could raise antitrust concerns and said the U.S. government could mediate the discussions or provide a narrow waiver for certain safety conversations.
Subscribers say slower progress means less value
The plaintiffs do not challenge a company’s decision to slow its own development, conduct its own safety testing or ask Congress and the White House for regulation. Their claim targets an agreement among rivals to limit the rate at which their products improve.
“AI will quickly spin out of human control and could kill us all if we allow AI safety and protocol... to be controlled by private self-serving agreements between the world’s most powerful ‘for profit’ technology companies,” Nick Rowley, the lead attorney for the plaintiffs, said in a statement cited by AP.
The complaint argues that subscribers paid for access to leading AI systems and ongoing product improvements. If competing companies jointly restrain that progress, the plaintiffs say, customers could pay the same prices for services that are less capable than they would have been under competition.
What the lawsuit asks the court to do
The proposed class action seeks damages and an injunction against agreements concerning the pace of AI development, model training, computing resources, release timing and the exchange of sensitive information used to enforce an alleged slowdown.
OpenAI has previously supported the idea of consistent federal safety requirements while rejecting the need to wait for an antitrust exemption. Altman said the company welcomed a federal framework but did not believe legislation or an exemption was required before work on safety could begin.
The plaintiffs’ position is narrower than a demand for unrestricted AI development. They say public authorities should set safety rules and that companies should not replace government oversight with private coordination among direct competitors.
A difficult test for antitrust law
The case presents an unusual question: whether public agreement with a safety proposal can establish an unlawful business arrangement when the companies have not announced a specific price, production quota or product restriction.
Sen. Josh Hawley, a Republican from Missouri, has already rejected the idea of granting the companies an antitrust exemption for joint safety work. He said there was “no world” in which he would support giving the largest technology companies such permission to collaborate without restrictions.
The court will first have to decide whether the consumers can pursue the case as a class and whether their complaint plausibly alleges an agreement that harmed them. For now, the filing places the companies’ public calls for coordinated restraint directly inside an antitrust dispute over who gets to decide how quickly commercial AI products should improve.