The full picture
A proposed class action filed September 18, 2026, in the U.S. District Court for the Northern District of California (Buist et al. v. Anthropic PBC et al., No. 3:26-cv-10693) accuses Anthropic, OpenAI, SpaceXAI LLC, and Google of violating Section 1 of the Sherman Act by agreeing to slow the pace of frontier AI development. The four named plaintiffs are paid subscribers to ChatGPT, Claude, Grok, and Gemini, suing on behalf of a proposed nationwide class. The complaint is pled as an unlawful-per-se restraint and, in the alternative, under quick-look and rule-of-reason analysis, seeking treble damages under the Clayton Act and an injunction. The filing states explicitly that plaintiffs are not challenging unilateral safety testing, outside evaluations, environmental safeguards, or government regulation.
How it developed
25 September 2026September 25 reporting confirmed the lawsuit's docket as Buist et al. v.
Anthropic PBC et al. (No. 3:26-cv-10693) and identified SpaceXAI LLC as the specific entity through which Musk controls the Grok chatbot business. Coverage of the suit, which accuses Anthropic, OpenAI, SpaceXAI, and Google of agreeing after a round of CEO statements on September 12 to slow AI development, has extended into wider debates over whether AI governance should be set by industry coordination or government.
24 September 2026Report that a near-finalized formal agreement between OpenAI and Anthropic for systematic mutual alignment testing was apparently stopped, with antitrust concerns cited speculatively
22 September 2026A September 22 analysis called the suit, which accuses Anthropic, OpenAI, SpaceXAI, and Google of agreeing to slow AI development, legally thin, noting it cites no private communications and that three of the four plaintiffs are lawyers.
The analysis flagged a contradiction: lead attorney Nick Rowley warns of existential risk from uncontrolled AI while the plaintiffs simultaneously advocate for unconstrained AI development. A separate analysis warned the suit may push labs to abandon joint safety statements in favor of unilateral policies.
First in the edition · 22 September 2026
21 September 2026Eric Schmidt told The Economist that an AI pause is not realistic given US competitive incentives, framing development as a US-China race
20 September 2026Four paying subscribers to ChatGPT, Claude, Grok, and Gemini filed a federal antitrust class action on September 18, 2026, in the Northern District of California against Anthropic PBC, OpenAI Opco LLC, SpaceXAI LLC, and Google LLC, alleging the companies illegally agreed to slow AI model development.
The complaint centers on September 12, when Anthropic CEO Dario Amodei published an essay urging industrywide AI deceleration and OpenAI CEO Sam Altman, SpaceXAI CEO Elon Musk, and Google DeepMind co-founder Demis Hassabis each publicly agreed the same day. Plaintiffs frame capability improvements as the product subscribers pay for, seek treble damages and an injunction, and explicitly exclude unilateral safety testing and government regulation from their claims. None of the four defendants had responded publicly at the time of filing.
September 26 reporting on a September 18 antitrust suit alleging Anthropic, OpenAI, SpaceXAI, and Google coordinated to slow AI development filled in the complaint's evidentiary core: Amodei's essay acknowledged that coordinating on pace would need a government antitrust waiver, Altman responded publicly that OpenAI would not wait for one, and OpenAI's policy chief confirmed weeks of prior collaboration, a sequence the complaint treats as proof of antitrust awareness and intent.
Senator Hawley had rejected Amodei's proposal at a Senate hearing before the filing.
18 September 2026Bloomberg Law, AI Weekly, and other outlets report on the lawsuit; lead attorney Nick Rowley makes public statements on AI governance.
17 September 2026Vinson & Elkins publishes analysis outlining four antitrust-compliant pathways for AI lab coordination.
10 August 2026Signatories of an AI pacing letter, including Dean W. Ball, clarify they advocate only a temporary slowdown to a rate still faster than current AI progress rates.