A class action lawsuit filed in a federal court in California accuses Anthropic, OpenAI, SpaceXAI —the artificial intelligence company of Elon Musk— and Google of having coordinated a supposed agreement to slow down the development of their artificial intelligence systems.
The case, filed on September 18 in the U.S. District Court for the Northern District of California, is promoted by four consumers who pay subscriptions to ChatGPT, Claude, Grok, or Gemini and who intend to represent a national class of users. The lawsuit claims that the coordination among the companies could violate U.S. antitrust legislation by limiting the pace at which competing products improve.
The lawsuit focuses on the call to "slow down" AI
The plaintiffs place the supposed agreement around the open debate during September about the need to reduce the pace of development of the most advanced models.
On September 12, the CEO of Anthropic, Dario Amodei, published a text in which he defended industry coordination to "modulate" the pace of AI development and reinforce safety measures. On the same day, Sam Altman, Elon Musk, and Demis Hassabis publicly expressed their support for different parts of the proposal.
The lawsuit interprets those statements as coordination among competitors that could reduce the pace of improvement of their products. The plaintiffs argue that, without that agreement, a company that decided to unilaterally slow down its development would risk losing users, revenue, and competitive position against the rest.
What the case would have to prove
The procedure is raised under the Section 1 of the Sherman Act, which prohibits certain agreements between companies that restrict competition. The central issue will be to determine whether there was really an agreement among the companies and whether that coordination had relevant anticompetitive effects.
The lawsuit claims that agreeing on the speed of improvement of competing products would amount to a restriction of supply or innovation. However, the filing of a lawsuit does not imply that those accusations have been proven or that the court has determined that the companies acted contrary to antitrust legislation.
Consumers request, among other measures, that the action be recognized as a class action, a declaration that an infringement of antitrust regulations occurred, and a court order that prevents certain forms of coordination between competitors.
The companies have not publicly responded to the lawsuit
Anthropic, OpenAI, Google, and SpaceXAI had not publicly responded to requests for comments on the lawsuit at the time of the latest published information.
However, there is a prior position from Sam Altman on the debate. The CEO of OpenAI stated that the company supported a federal framework that established common safety requirements for AI developers, but argued that it was not necessary to wait for an exemption from antitrust rules to begin working on shared safety measures. He also emphasized that "pacing" —modulating the pace of development— does not mean stopping progress.
The debate on a possible antitrust exemption
The legal dimension of the lawsuit takes on special relevance because Amodei himself had warned about the possible competition problems that could arise from coordination between rival companies. In his proposal, he suggested that the U.S. government could mediate or facilitate certain conversations between laboratories and considered a limited exemption for some issues related to safety.
The case thus raises an open question: how far can competing companies go in establishing common safety standards without that coordination conflicting with antitrust rules. For now, the litigation is in an initial phase and there is no judicial resolution on the substance of the accusations.