Behind the AI slowdown, is there a "monopoly conspiracy"?

On September 12, Anthropic CEO Dario Amodei published an article calling for industry-wide cooperation to slow down the pace of AI advancement in exchange for stronger safety measures.

That same day, OpenAI's Sam Altman, SpaceXAI's Elon Musk, and Google DeepMind's Demis Hassabis publicly voiced their support.

Within seven days, this industry consensus was breached from two directions. Last Friday, four companies were sued in the U.S. District Court for the Northern District of California on charges of jointly restraining competition. The next day, the President of the United States called the same matter a "scam" on social media.

What does the complaint say?

The plaintiffs are four individuals, all paying subscribers of ChatGPT, Claude, Grok, or Gemini. Lead attorney Nick Rowley is seeking class-action status on their behalf, representing all paying subscribers of these services across the United States.

The allegation itself is plain, even unexpectedly pragmatic: these four companies coordinated to slow down development, diminishing the value consumers received for their money.

The complaint argues that the coordination primarily occurred on September 12. Anthropic CEO Dario published his article, and the other three responded the same day. The complaint also asserts that this coordination had been taking shape for months prior.

The hardest part of this lawsuit to refute is not in the plaintiffs' arguments — it's in the defendants' own articles.

In his article calling for deceleration, Dario himself raised antitrust concerns. His approach was to make a request of the government: "I hope the government can grant a narrow exemption for certain types of safety conversations."

That sentence has now become the fulcrum of the complaint.

Because it simultaneously proves two things: First, he knew what industry-coordinated deceleration means legally. Second, he didn't get that exemption.

And Altman's response pushed the issue even further. He said OpenAI welcomes "a federal framework that sets consistent safety requirements," but immediately added: "We don't think we need to wait for an antitrust exemption or legislation to start doing the work of providing that confidence."

Translated: We know there's no exemption, and we intend to proceed anyway. In an antitrust lawsuit, statements like these typically don't help the defendants.

Trump Fires Again

On Saturday, Trump announced on Truth Social that he would establish an "AI Force" modeled after the Space Force, and would soon appoint a new AI czar. He added a job posting note to the position: "Only high-IQ individuals need apply." He did not name a pick.

The position previously belonged to David Sacks, Trump's AI and crypto czar, who stepped down earlier this year due to term limits for special government employees and transitioned to an advisory role. He is reportedly expected to attend Trump's state dinner with Xi Jinping.

Trump left no ambiguity about his commitment to the industry: "We will never in any way impede or kill the growth of this great industry. Instead, we will cherish it, help it, and watch over its growth!"

As for the calls for slowing down, his characterization was one word: a scam.

His stated reason was competition: America must lead, must outrun China. The line he repeated was "whoever wins AI wins." As for what the AI Force actually is — its budget, its structure, its headcount, when it would be built — the White House said not a single word.

On the surface, the courtroom and the White House stand on opposite sides. But put them together, and they point to the same conclusion. An AI company that wants to tap the brakes on safety theoretically has three paths:

The first: brake on your own. Slow down unilaterally while rivals don't. That is handing over market share directly. So in reality, no company would do this alone.

The second: brake together. Industry coordination, everyone slows down. That is exactly what happened on September 12, and exactly what is now being sued over.

The third: be required to brake. Congress legislates, setting uniform standards everyone must follow. This is the "federal framework" Altman spoke of, and the only path he considers safe under antitrust law, because acting as legally required does not constitute collusion.

And this weekend's two news items happened to block the second path and veto the third.

The lawsuit turned industry coordination into legal liability. The president's statement ensured federal legislation will not arrive in the foreseeable future. He made it clear he will not impede the industry in any way, and that safety concerns are a scam.

Only the first path remains. And the first path, no company will take.

This lawsuit is currently just a complaint. It has not yet undergone any substantive review by any court, and the four defendants have not yet responded. Whether the allegations hold up will not be answered for a long time.

But it has already accomplished something that does not require winning: it has changed the nature of the term "responsible AI." For the past three years, it was a public relations asset. Since last Friday, it is also potentially material that could be used as evidence. How this will affect the next CEO who wants to publicly call for slowing down probably does not need to wait for a verdict.

In the same week, AI researchers including Nobel laureate Geoffrey Hinton warned that Congress has roughly one year left to regulate the industry.

Welcome to join the official BlockBeats community:

Telegram Subscription Group: https://t.me/theblockbeats

Telegram Discussion Group: https://t.me/BlockBeats_App

Official Twitter Account: https://twitter.com/BlockBeatsAsia