Skip to main content

Antitrust Suit Says the AI Slowdown Call Was an Illegal Pact

A federal class action filed Friday alleges that Anthropic, OpenAI, xAI, and Google made an illegal agreement to slow the pace of their AI development. The suit centers on Dario Amodei's September 12 essay calling for industry-wide coordination, followed within hours by endorsements from Sam Altman and Elon Musk, and claims the coordination harmed paying subscribers.

On this page

The suit, and what it targets

A federal antitrust class action filed on Friday alleges that Anthropic, OpenAI, xAI, and Google made an illegal deal to collectively slow the pace of their AI development. The complaint was filed by four consumers, who claim the coordination between supposedly competing labs harmed paying subscribers of products like ChatGPT, Claude, Gemini, and Grok. The defendants are Anthropic PBC, OpenAI OPCO LLC, xAI, and Google LLC.

At the center of the complaint is Dario Amodei's September 12 essay, "We Must Pace the Frontier," which called for industry-wide coordination to slow AI progress. Within hours, Sam Altman had publicly endorsed the call, saying "I agree with Dario that we need to pace the frontier," and Elon Musk and Google executives reportedly followed. The plaintiffs argue these public statements, combined with a July 2026 industry working group and OpenAI's "An Alien Mind" essay from September 6, amount to an unlawful horizontal agreement between competitors.

Allegations, not findings

Every claim in the suit is just that: a claim in a freshly filed complaint, and none of the companies has been found liable for anything. Expect the defendants to argue that whatever coordination occurred was safety commitments rather than market allocation, and that pacing language in essays is not a pricing agreement. Antitrust cases targeting speech about coordinated conduct face a high bar, and this one will be tested early.

The political framing cuts both ways too. The slowdown debate began when Amodei asked for permission to slow down, Altman and Musk backed him, and President Trump rejected the whole idea the next day. If the industry's safety coordination is now also an antitrust theory, every CEO who endorses a rival's caution post may need a lawyer in the room.

Why the timing is awkward for everyone

The suit lands in the middle of the busiest governance fortnight of the year: a Senate document demand to OpenAI with an October 1 deadline, the first US-China AI safety dialogue with its proposed notification mechanism, a UN panel warning that traditional safeguarding is unraveling, and a royal summit in Scotland. The industry asked Washington for pacing permission; Washington's antitrust response, effectively, is that pacing itself may be the violation.

For paying subscribers, the plaintiffs' theory is straightforward: coordination between competitors on how slowly to improve products is the kind of agreement antitrust law exists to police. Whether safety coordination between rivals is different in kind is the question the court will now have to answer, and the answer will shape what labs can say to each other about pacing for years.

CuriousLM runs supported AI models locally on your device. Try CuriousLM.