Four AI subscribers accuse Anthropic, OpenAI, SpaceXAI and Google of illegally coordinating a development slowdown that would leave paying customers with less value.
The proposed nationwide class action was filed Friday in the U.S. District Court for Northern California. It covers subscribers to ChatGPT, Claude, Grok and Gemini.
The lawsuit alleges that the four companies violated antitrust laws by coordinating how quickly they would develop artificial intelligence. Those laws generally bar competitors from making agreements that restrain competition.
The plaintiffs do not argue that companies must ignore safety. They contend each company can adopt safeguards on its own, but rivals cannot collectively restrain development and reduce the value of paid subscriptions.
The complaint’s lawyers stated their position plainly, as CBS News reported:
“The antitrust laws do not permit competitors to decide among themselves that competition is too dangerous,”
The alleged coordination centers on a September 12 essay from Anthropic CEO Dario Amodei. He urged AI companies to slow advances across the industry while adding stronger safety measures.
Amodei described the goal as “pacing the frontier.” His three-point plan also called for the federal government to mediate “or at least enable” discussions among competing AI labs.
He wrote that Washington would need to “issue a narrow waiver for certain kinds of safety conversations.” Such a waiver would address the antitrust risk created when direct competitors discuss coordinated limits.
The lawsuit says OpenAI CEO Sam Altman, Elon Musk of SpaceXAI and Demis Hassabis of Google DeepMind confirmed their agreement with the proposed slowdown that same day. Their exact words were not provided.
The Associated Press reported that public calls for collective action created a direct tension between safety cooperation and competition law. Industry leaders have discussed government mediation, regulation or an antitrust exemption.
That distinction matters. A company may decide that its own product needs more testing. The legal dispute begins when competitors allegedly agree that all of them should move more slowly.
The four plaintiffs pay for access to one or more leading AI services. They seek to represent a nationwide class of customers who subscribe to ChatGPT, Claude, Grok or Gemini.
No specific damages figure was reported. The complaint argues that coordinated restraint would leave consumers receiving less value than open competition otherwise would produce.
Breitbart reported that the plaintiffs view an agreement among the leading AI rivals to move slower than competition would produce as harmful to consumers.
The companies had not responded to requests for comment as of Saturday. That leaves the central allegation unanswered: whether public support for an industrywide slowdown became an actual agreement among competitors.
Amodei has warned that poorly developed artificial intelligence could create severe risks. In an interview with CBS senior business and technology correspondent Jo Ling Kent, he acknowledged the speed of recent progress.
Amodei said, “I don't think I fully just appreciated what it would actually be like when the progress was as fast as it was.”
He also drew a sharp line between responsible and irresponsible development. “If we build in the right way, I think the probability of something bad happening is very low,” he said. “If we build in the wrong way, the probability of something bad happening is very high.”
Amodei separately warned that swarms of rogue AI agents could take over the internet in as little as six months. The lawsuit does not dismiss such concerns; it challenges who gets to set the pace and by what legal authority.
Key details remain unresolved, including the precise terms of any alleged agreement, the relief sought and the complete wording of the rival executives’ responses. The complaint’s case will depend on whether those statements show shared concern or unlawful coordination.
Safety cannot become a private license to weaken competition. If powerful companies want special permission to coordinate, they should seek it openly and answer to the law.