In a proposed class action in the Northern District of California, four AI chatbot subscribers allege that Anthropic, OpenAI, Google, and SpaceXAI unlawfully agreed to slow AI development. SpaceXAI has moved to sever and transfer the claims against it to the Northern District of Texas or compel arbitration, and the plaintiffs opposed the motion on October 27, 2026. A hearing before Magistrate Judge Nathanael M. Cousins is set for November 18, 2026.
Latest development Oct 27, 2026
On October 27, 2026, the plaintiffs filed a brief opposing SpaceXAI’s motion to sever and transfer the claims against it to the Northern District of Texas or to compel arbitration. They argue that the arbitration notice on SpaceXAI’s sign-up screen appeared in small, muted text and was insufficient to bind users. A hearing before Magistrate Judge Nathanael M. Cousins is set for November 18, 2026.
Why it matters
The case will test whether public calls among rival AI labs for coordinated safety “slowdowns” can amount to per se unlawful collusion under the Sherman Act. The outcome affects four major AI developers and any future industry coordination on the pace of AI development.
Background
On September 18, 2026, four paying subscribers to AI chatbots sued Anthropic PBC, OpenAI, Inc., SpaceXAI LLC, and Google LLC in a proposed class action. They filed in the San Francisco Division of the U.S. District Court for the Northern District of California (Case No. 3:26-cv-10693). SpaceXAI is the entity through which Elon Musk now controls the former X.AI business. The named plaintiffs are Florida attorneys Charles Buist and Nick Spetsas and California residents Cheyenne Hunt and Christine Bullock. Each pays for a subscription to one of the defendants’ main consumer products: ChatGPT, Claude, Grok, or Gemini.
The complaint alleges that the four companies violated Section 1 of the Sherman Act by agreeing to slow the pace at which their competing AI products improve. The plaintiffs trace the alleged agreement to an essay that Anthropic chief executive Dario Amodei published on September 12, 2026. The essay called for industry-wide coordination to “pace the frontier” on safety grounds. Within hours, Elon Musk, OpenAI chief executive Sam Altman, and Google DeepMind co-founder Demis Hassabis made public statements of endorsement or qualified agreement. The complaint pleads the claim as a per se unlawful restraint and, in the alternative, under quick-look and rule-of-reason analysis.
The plaintiffs say they do not object to any single company slowing its own development. They argue that antitrust law forbids competitors from replacing independent decisions about product improvement with collective agreement, which they characterize as an unlawful restriction of output. They seek class certification, treble damages, and a declaration that the conduct violated federal antitrust law, and they have demanded a jury trial. Under Section 16 of the Clayton Act, they also seek an injunction barring the defendants from coordinating on development pace, training compute, release timing, or capability checkpoints. The injunction would also bar the information exchange needed to monitor any such agreement.
On September 21, 2026, the court issued summonses to the four defendants. OpenAI was served on September 23, 2026, and Anthropic, Google, and SpaceXAI on September 24, 2026, with answers due on October 14 and October 15, 2026, respectively. On October 13, 2026, SpaceXAI moved to sever and transfer the claims against it to the Northern District of Texas or to compel arbitration. The plaintiffs opposed the motion on October 27, 2026, arguing that SpaceXAI’s sign-up screen displayed its arbitration notice in small, muted text that could not bind users. A hearing before Magistrate Judge Nathanael M. Cousins is set for November 18, 2026, and no defendant has answered or moved to dismiss the complaint.
Timeline 9 events
-
FilingPlaintiffs oppose SpaceXAI’s arbitration and transfer motion
Plaintiffs filed a brief opposing SpaceXAI’s motion, arguing the arbitration notice on SpaceXAI’s sign-up screen was displayed in small, muted text and was insufficient to bind users to arbitration.
AI Users Say SpaceXAI Can't Split, Arbitrate Slowdown SuitPlaintiffs say SpaceXAI must face US antitrust claims in court
-
FilingSpaceXAI moves to sever, transfer, or compel arbitration
SpaceXAI asked the court to sever its part of the case and transfer it to the Northern District of Texas. In the alternative, it asked the court to compel arbitration of the plaintiffs’ claims against it.
AI Slowdown Pact Suit: Four Labs, One Sherman Act ClaimAI Users Say SpaceXAI Can't Split, Arbitrate Slowdown Suit
-
OrderGoogle and SpaceXAI are served with the summons
Google LLC and SpaceXAI LLC were each served with the summons and complaint, with answers due October 15, 2026.
-
OrderAnthropic is served with the summons
The summons was returned executed as to Anthropic on September 24, 2026. Anthropic’s answer is due on October 15, 2026.
-
OrderOpenAI is served with the summons
OpenAI OpCo, LLC was served with the summons and complaint, with its answer due October 14, 2026.
-
OrderCourt issues summonses to all four defendants
-
FilingSubscribers file antitrust class action in Northern District of California
Four paying AI subscribers filed Buist et al. v. Anthropic PBC et al., alleging a Sherman Act Section 1 conspiracy to slow AI development.
OpenAI, Anthropic, Google, SpaceXAI Hit With Antitrust LawsuitAnthropic, OpenAI, Google and SpaceXAI face lawsuit claiming AI slowdown harms subscribers and violates antitrust laws
-
OtherRival CEOs publicly endorse or respond to Amodei’s proposal
Elon Musk, Sam Altman, and Demis Hassabis reacted publicly to Amodei’s essay within hours. The complaint cites these statements as evidence of a coordinated agreement.
OpenAI, Anthropic, Google, SpaceXAI Hit With Antitrust LawsuitLawsuit says Anthropic, OpenAI, SpaceXAI, Google made illegal AI deal
-
OtherAnthropic CEO publishes essay urging industry-wide AI deceleration
Dario Amodei called for coordinated pacing of frontier AI development among labs on safety grounds. The complaint identifies the essay as the origin of the alleged agreement.
Lawsuit says Anthropic, OpenAI, SpaceXAI and Google made illegal deal on AI slowdownLawsuit Alleges Anthropic, OpenAI, SpaceXAI and Google Made Illegal Agreement for AI 'Pacing'
Experimental AI forecast
How it is madeHow will the court resolve SpaceXAI’s motion to sever and transfer the claims against it to the Northern District of Texas or compel arbitration, by April 30, 2027?
- SpaceXAI claims leave the case30%
- Court denies SpaceXAI’s motion24%
- Still pending on April 30, 202722%
- Court splits SpaceXAI’s motion14%
- Plaintiffs drop claims against SpaceXAI10%
- Resolves by
- Apr 30, 2027in 6 months
- Record confidence
- Low
The most likely result, at 30%, is that SpaceXAI’s claims leave the Northern District of California by April 30, 2027, through arbitration or transfer. Courts usually enforce consumer arbitration and forum terms. The motion is fully briefed and set for hearing on November 18, 2026. The plaintiffs’ conspicuousness challenge and the fact that only one named plaintiff subscribes to Grok keep denial or a split ruling live.
How it resolves. Resolved from the N.D. Cal. docket in Case No. 3:26-cv-10693. “SpaceXAI claims leave the case” if a court order compels arbitration of, transfers, or dismisses all claims against SpaceXAI. “Court denies SpaceXAI’s motion” if the motion is denied in full and all claims against SpaceXAI stay in N.D. Cal. “Court splits SpaceXAI’s motion” if some plaintiffs’ or claims’ claims against SpaceXAI leave the case and others remain. “Plaintiffs drop claims against SpaceXAI” if plaintiffs voluntarily dismiss SpaceXAI or the whole case before any ruling. “Still pending” if none of these has occurred by April 30, 2027.
Reasoning
Starting point. Federal motions to compel arbitration or enforce forum-selection clauses in Ninth Circuit consumer online-terms cases: Roughly 55-65% of such motions are granted at least in part, but sign-in-wrap cases with inconspicuous notice are denied more often under Berman v. Freedom Financial (9th Cir. 2022).
Courts grant most motions to compel arbitration or enforce forum-selection clauses in consumer online-terms cases, often around 55-65% at least in part. Ninth Circuit law under Berman requires reasonably conspicuous notice, and the plaintiffs attack SpaceXAI’s sign-up screen as small, muted text, which lowers that rate here. Only one named plaintiff pays for Grok, so SpaceXAI’s terms may bind only that plaintiff, which makes a split ruling plausible. Atlantic Marine favors enforcing valid forum clauses, which supports transfer if the terms bind the plaintiffs. The November 18, 2026, hearing makes a ruling within five months likely, but the court may defer while the other defendants respond or while magistrate-judge consent is resolved, which supports a meaningful pending share. Plaintiffs could also drop SpaceXAI to keep the case consolidated in California, given the early posture. The leading alternative is outright denial at 24%. The previous forecast found the evidence insufficient, so its numbers are not compared.
Conspicuousness of SpaceXAI’s sign-up arbitration notice under Berman · Only one named plaintiff subscribes to Grok · Strong federal policy enforcing arbitration and forum clauses · Unresolved magistrate-judge consent and other defendants’ pending responses · Plaintiffs’ incentive to keep all defendants in one forum
Generated automatically by Claude Opus 5.5 on Nov 1, 2026 from this case’s record on this site, and published without editorial review. The probabilities are the model’s judgment, not measured frequencies, and they are scored publicly as cases resolve. Not legal or investment advice. How the forecasts are made
Disclosure: Anthropic, which makes the model that wrote this forecast, is a party to this case.