Subscribers sue AI giants: Did they agree to slow development?

Four paying users of AI chatbots have filed a federal lawsuit in the US against Anthropic, OpenAI, SpaceXAI and Google. The case has been filed in the federal court for the Northern District of California. The users subscribe to Claude, ChatGPT, Grok or Gemini and want to represent all paying users in the US.

They believe the companies have violated antitrust law (Sherman Act, Section 1). The law prohibits competitors from making agreements that restrict the market.

What happened?

On September 12, 2026, Anthropic CEO Dario Amodei published an essay. In it, he argued that AI labs should jointly set limits on how fast AI development may proceed. Within a few hours, Elon Musk (SpaceXAI), Sam Altman (OpenAI) and Demis Hassabis (Google DeepMind) expressed their support.

The plaintiffs also point to a statement from July 2026 signed by top executives from several AI companies. It mentions strong competitive pressure that makes it hard for one company to slow down on its own. According to the plaintiffs, this suggests the cooperation began before September.

Why do the plaintiffs think it is illegal?

The core of the complaint is that the companies would otherwise compete to make the best products. If they agree to move more slowly than each of them would have chosen alone, customers get worse products for their money. The plaintiffs describe it as restricting supply.

Attorney Nick Rowley, who is leading the case, has said according to the news agency AP that AI could quickly slip out of human control and in the worst case could kill us if safety is decided through "self-serving agreements." So the plaintiffs do not want slower development. They want safety questions not to be decided by the companies themselves.

What does it mean for you?

The case has only just begun, and no ruling has been made. None of the defendants had immediately commented on the case when it was reported. It has also not been decided whether a public call for common rules even counts as an illegal agreement.

The case reveals a tension that could matter for ordinary users. Many want safe AI, but can the companies cooperate on safety without violating antitrust law? The answer could affect how AI tools are developed and regulated going forward.