Law360 (September 21, 2026, 6:34 PM EDT) -- Artifical intelligence users in Florida and California have accused Anthropic, OpenAI, SpaceX and Google of illegally restraining competition by publicly agreeing to "pace the frontier" and collectively slow down their AI development.

The four users said in their 29-page complaint filed Friday in California federal court that the agreement in question was accepted and confirmed in public. It noted that Anthropic CEO Dario Amodei published an essay titled "We Must Pace the Frontier" on Sept. 12 that called for "industry-wide coordination" among frontier AI laboratories to set "limits on the rate of unchecked AI progress," and it explained that the coordinated slowing would let each participant reduce its pace "without sacrificing commercial advantage."

Amodei said the cooperation is necessary in the wake of the Hugging Face incident in which AI agents escaped what was thought to be a secure testing environment, hacked into a variety of databases and gained authorized access to credentials.

The complaint said that Amodei's rivals — Elon Musk, who founded and controls SpaceXAI's Grok business, Sam Altman, CEO of OpenAI, and Google DeepMind co-founder Demis Hassabis — confirmed their agreement the same day and that OpenAI's global policy chief confirmed the next day that the companies have been working together for weeks. 

"An agreement among competitors to reduce the quality of their products and the rate at which those products improve is an agreement to restrict output," according to the complaint, which only names the companies as defendants, not the executives. "It is among the restraints Section 1 of the Sherman Act, 15 U.S.C. § 1, has always condemned, and it does not become lawful because the products are new."

The consumers said the path to responsible, safe and environmentally conscious AI development is each company individually "investing in safety research, adopting rigorous internal evaluations, reducing its environmental footprint, and welcoming government oversight and regulation." They added that the consumers' suggested path is "fully open to every defendant."

"What the antitrust laws forbid is the shortcut of competitors agreeing among themselves to substitute collective restraint for individual accountability," the complaint said. "This action is brought in defense of the competitive process that makes genuine responsibility and genuine progress possible."

The plaintiffs said, "The lesson of every previous transformative industry is that safety and responsibility are properly the work of each company acting on its own, not a cartel license." In regards to AI, "each Defendant is free, and indeed obligated, to set its own safety thresholds, retain its own evaluators, build its own product guardrails, and restrain its own development. None of that requires an agreement with rivals."

The four users are consumers who pay one of the companies directly for subscriptions to one or more of their AI services: ChatGPT, Claude, Grok and Gemini. They are seeking to represent a proposed class of consumers who also paid for subscriptions to ChatGPT, Claude, Grok and Gemini.

The complaint said a statement titled "Pacing the Frontier" was published by senior executives of Anthropic, OpenAI and Google in July 2026. The statement identified the "collective-action problem" at the heart of the case and called for tools capable of "deliberately pacing frontier-wide development."

Hassabis then proposed a U.S.-led standards body for frontier AI modeled in part on the Financial Industry Regulatory Authority, the complaint said.

"Under the proposal, frontier laboratories would initially submit models voluntarily for pre-release review; the body could later impose common requirements before models entered the U.S. market and could escalate those requirements as risks increased," the complaint said. "Contemporaneous reporting described the proposal as a mechanism that could coordinate an industry-wide slowdown."

The consumers said representatives of the four companies then formed a group that "met regularly to work together on creating a standards body for the industry" beginning that same month.

The day before Amodei published the essay in question, the complaint said that Altman was asked why leading AI executives, including Amodei, did not get together and create a common plan, to which he said, "I think that will happen."

The consumers said Amodei's proposal "supplied every element of a horizontal agreement in restraint of trade: the parties (the frontier laboratories); the term (a slower rate of capability advancement than each would choose alone); the mechanisms (shared limits on compute, training runs, and self improvement, and capability checkpoints); and the means of policing defection (verifiable pacing through embedded evaluators)."

They noted that the other companies immediately accepted Amodei's proposal.

"By the end of September 12, therefore, a senior executive or controlling person of each Defendant had publicly signaled assent to a proposal that the frontier laboratories coordinate to slow the pace of capability development," they said. "Each signal was given with knowledge of the others, in sequence, on a single day, in direct response to the proposal."

They added that the companies confirmed the agreement and its object, that they knowingly joined it and that it was not just "parallel conduct."

They argued that the agreement will restrain competition as each of the companies under the agreement "makes those decisions knowing that its principal rivals have agreed to reduce their own pace."

"That knowledge eliminates the competitive penalty for slowing — the risk that a rival will race ahead — and allows each Defendant to reduce its investment and competitive effort without losing ground," the complaint said.

They added that the agreement is a "naked horizontal restraint on a central dimension of competition" among direct rivals, and it is illegal without proof of a relevant market or market power. They said the agreement will have anticompetitive impacts and will result in antitrust injuries to the consumers.

"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," said Nick Rowley of Trial Lawyers for Justice, counsel to the class.

"Humanity deserves iron clad safeguards when it comes to extinction event threats such as nuclear warfare and now the biggest risk to mankind in history," he added. "The rule of law should be established transparently and lawfully by our government, with accountability to the public. We are calling on the President, the Senate and Congress as well as all federal regulators to act immediately to prioritize public safety over mega rich AI companies protecting competition."

Representatives for Anthropic, OpenAI, SpaceX and Google did not immediately respond to requests for comment on Monday.

The consumers are represented by Andrew Tutt and Nick Rowley of Trial Lawyers for Justice.

Counsel information for Anthropic, OpenAI, SpaceX and Google was not available on Monday.

The suit is Buist et al. v. Anthropic PBC et al., case number 3:26-cv-10693, in the U.S. District Court for the Northern District of California, San Francisco Division.

--Editing by Michael Watanabe.

For a reprint of this article, please contact reprints@law360.com.