Four consumers filed a federal class-action complaint in San Francisco on September 18, 2026, accusing Anthropic PBC, OpenAI OpCo LLC, SpaceXAI LLC, and Google LLC of agreeing among themselves to slow the pace at which their competing frontier AI products improve — a deal the plaintiffs say violates Section 1 of the Sherman Antitrust Act.

Bloomberg Law reported the suit Friday night. Unite.AI, republished via TradePoint, published a detailed complaint summary Saturday. The case is Buist v. Anthropic PBC, U.S. District Court for the Northern District of California, San Francisco Division, No. 3:26-cv-10693. None of the defendants immediately responded to Bloomberg Law’s request for comment.

Insanity Developing previously covered Anthropic CEO Dario Amodei’s September 12 “pace the frontier” essay and Washington’s resistance to an Anthropic bid for an AI antitrust safety waiver. This filing is the consumer lawsuit that followed those public exchanges — not a rehash of the essay alone.

Who sued and what they claim

The named plaintiffs are Charles Buist and Nick Spetsas of Florida and Cheyenne Hunt and Christine Bullock of California. They sue individually and on behalf of a proposed nationwide class of U.S. persons who, from September 12, 2026, onward, purchased a paid individual consumer subscription to ChatGPT, Claude, Grok, or Gemini directly from a defendant, with subclasses for each company’s subscribers.

Lead counsel is Nicholas C. Rowley, joined by Andrew T. Tutt, R. Stanton Jones, and Jakob Z. Norman of Trial Lawyers for Justice.

The complaint’s central theory is that an agreement among rivals to restrict the rate of product improvement is an output restraint. The plaintiffs plead the alleged restraint as unlawful per se and, in the alternative, under quick-look and rule-of-reason analysis. Their injury theory is an overcharge: paying the same subscription prices for products that improve more slowly than competition would otherwise produce.

The filing defines a U.S. market for paid consumer subscriptions to general-purpose frontier generative-AI assistants — covering paid tiers of ChatGPT, Claude, Grok, and Gemini — and alleges on information and belief that the four defendants account for at least 80 percent of paid consumer subscriptions in that market. It notes ChatGPT Plus at roughly $20 per month.

“Plaintiffs challenge only what the antitrust laws forbid: an agreement among competitors about how fast their competing products will improve,” the complaint states, according to Bloomberg Law. “Congress has granted no exemption for that agreement.”

Timeline in the complaint

According to Unite.AI’s summary of the filing, the alleged coordination began months before the September 12 public exchange:

  • July 2026: Anthropic, OpenAI, and Google representatives formed a working group on an industry standards body; on July 14, Google DeepMind co-founder Demis Hassabis publicly proposed a U.S.-led standards body modeled in part on FINRA.
  • September 6: OpenAI chief scientist Jakub Pachocki’s essay “An Alien Mind” described coordination among frontier developers to slow future development as one principal option.
  • September 10: WIRED reported OpenAI had sought congressional guidance on whether coordinating an industry-wide slowdown could raise antitrust risk.
  • September 11: In a Fortune interview, OpenAI CEO Sam Altman said he thought a common industry plan would happen, without detailing private discussions.
  • September 12: Amodei published “We Must Pace the Frontier,” writing: “We must slow the pace at which we improve the capabilities of AI models.” The filing alleges Musk endorsed the proposal within about an hour; Altman said he agreed and committed OpenAI to the plan’s first step; Hassabis called it “the right path forward.”
  • September 14: Altman said progress would proceed more slowly and that OpenAI would not wait for an antitrust exemption before working with industry colleagues.
  • September 15: OpenAI Global Policy Chief Chris Lehane confirmed weeks of work with Anthropic and Google DeepMind.

The complaint also points to a July 2026 “Pacing the Frontier” employee statement (1,386 signatories, with organizational support from Guidelight AI Standards and Encode AI) that described competitive pressure not to slow unilaterally and asked the U.S. government to support tools to deliberately pace frontier development.

What plaintiffs do — and don’t — challenge

The plaintiffs seek treble damages under the Clayton Act and an injunction barring horizontal agreements on the rate of AI development or improvement, limits on training compute or training runs, coordinated release delays, capability checkpoints, or exchanges of competitively sensitive information used to police such a restraint.

They expressly do not challenge unilateral safety measures, independent slowing, lawful safety research, compliance with government requirements, petitioning government, or legitimate standard-setting. In the filing, they say they take AI safety seriously but believe guardrails should come through public regulation and juries rather than private coordination among rivals.

Whether public endorsements and industry working groups amount to a forbidden agreement — or protected speech and standard-setting — is now for a federal court to sort out. For paying subscribers, the immediate question the suit poses is simpler: if rivals really did agree to improve more slowly together, did customers keep paying full price for a slower product?

Sources