Subscribers sue Anthropic, OpenAI, Google and SpaceXAI over alleged pacing pact

takeaways:

Four individuals who pay for AI subscriptions filed a proposed class action on 18 September 2026 against Anthropic, OpenAI, Google and SpaceXAI, alleging the companies illegally agreed to slow the pace of AI development. The lawsuit, Buist et al. v. Anthropic PBC et al., was lodged in the US District Court for the Northern District of California.

The plaintiffs said the case centres on a 12 September essay by Anthropic chief executive Dario Amodei, titled "We Must Pace the Frontier", which called for industry-wide slowing of AI capability gains. The complaint says OpenAI's Sam Altman, SpaceXAI's Elon Musk and Google DeepMind's Demis Hassabis signalled agreement with Amodei the same day.

Three of the plaintiffs pay for Claude, ChatGPT, Grok and Gemini; the fourth pays only for Claude, according to the complaint. They are seeking to represent AI subscribers nationwide. Their lead attorney is Nick Rowley.

The complaint also cites a July statement referencing "intense competitive pressure not to unilaterally slow" down, which the plaintiffs say points to prior awareness among the companies of the competitive stakes involved.

The plaintiffs argue that a coordinated slowdown in product improvement amounts to an unlawful restriction of output under Section 1 of the Sherman Act. They are seeking a liability finding and an injunction under Section 16 of the Clayton Act, according to ABC News.

The complaint states plaintiffs do not object to any single company slowing its own development for safety reasons, only to competitors agreeing to do so together. "The antitrust laws do not permit competitors to decide among themselves that competition is too dangerous," the complaint said.

The lawsuit comes after Congress recently stripped a related antitrust shield from a defence bill, according to Bloomberg Law. The Justice Department has signalled that narrower cybersecurity cooperation between AI labs would not raise the same antitrust concerns.

None of the four companies had responded to the lawsuit as of the filing. The case will test whether public coordination on AI safety pacing can survive scrutiny under US competition law.