OpenAI and Google Agreed to Slow AI Down, a Lawsuit Says

Four paying subscribers to ChatGPT, Claude, Grok and Gemini want a nationwide class action over the alleged deal.

Four paying chatbot customers sued Anthropic, OpenAI, SpaceXAI and Google on September 18, 2026, alleging the rivals agreed among themselves to slow AI development. The antitrust complaint was filed in the U.S. District Court for the Northern District of California, The Associated Press reported in an account ABC7 Los Angeles published.

The people on the hook here are subscribers. The four named plaintiffs pay for ChatGPT, Claude, Grok or Gemini, and they're asking to represent every other paid subscriber to those services in the country — which would include anyone in Los Angeles on a monthly plan. Their theory is that a slower race buys them less product for the same money.

What the complaint alleges, and what it doesn't

1. A one-day, public handshake. The suit points to September 12, 2026, when Anthropic chief executive Dario Amodei published an essay calling for industrywide cooperation on decelerating AI progress in favor of stronger safety work. OpenAI's Sam Altman, SpaceXAI's Elon Musk and Google DeepMind co-founder and chair Demis Hassabis each responded publicly the same day in agreement, the AP reported. The plaintiffs call that coordination; the companies haven't answered that characterization in court.
2. A slower start, months earlier. The complaint also cites a July 2026 statement signed by senior staff at several leading labs, which acknowledged the competitive pressure against slowing down alone and asked governments to back a global effort to slow automated AI development. The plaintiffs argue the coordination was taking shape by then. That's an allegation about intent, and it's the hardest kind to prove.
3. The legal claim. Rivals agreeing that their progress should run slower than competition would otherwise deliver hurts consumers, the plaintiffs argue — that's the antitrust violation as they frame it. No court has weighed it yet.
4. What is not alleged. The plaintiffs say they don't object to any single company deciding on its own to go slower for safety reasons, and they don't object to the industry asking Congress, the White House or an agency for regulation, or even for an antitrust exemption. What the law forbids, they argue in the complaint, is the shortcut of agreeing to "substitute collective restraint for individual accountability."
5. What the companies have said so far. Amodei's essay acknowledged the antitrust problem and suggested the federal government mediate the cross-lab conversations, or at minimum grant a narrow waiver covering safety discussions. Altman said on social media that OpenAI would welcome a federal framework with consistent safety requirements, but doesn't think the work needs to wait for an exemption or new legislation. Representatives for all four companies didn't immediately respond to the AP's request for comment on September 19, 2026, the outlet reported.

Nick Rowley, the lead attorney for the plaintiffs, told the AP that AI "will quickly spin out of human control and could kill us all" if safety protocols are set by private agreements among the largest for-profit technology companies.

The federal route the labs floated looks narrow. President Donald Trump has rejected the calls for regulation on social media and has called efforts to limit the technology a conspiracy, questioning why industry leaders would seek rules that he says would bankrupt them. On September 19, 2026, he said he's forming an AI task force and will appoint a czar for artificial intelligence, with little detail attached, per the same AP account. Sen. Josh Hawley, a Missouri Republican, said at a recent Senate hearing that he'd never vote to hand the most powerful companies in history an antitrust exemption.

Whether this becomes the nationwide class action the plaintiffs want is a judge's call.

Source: abc7.com, retrieved September 20, 2026.