Source: https://reinouttebrake.com/stories/ai-was-supposed-to-slow-down-now-theyre-being-sued-for-it - [Home](https://reinouttebrake.com/) - [AI](https://reinouttebrake.com/topic/ai) - AI Safety Antitrust Lawsuit: Can Rival Labs Slow Down Together? # AI Safety Antitrust Lawsuit: Can Rival Labs Slow Down Together? Subscribers accuse major AI labs of coordinating slower development. Explore the allegations and the tension between competition and collective safety measures. 20 Sept 2026 2 min read By [Reinout te Brake](https://reinouttebrake.com/about) Updated 17 Oct 2026 ![The United States Supreme Court building in Washington, D.C.](https://reinouttebrake.com/assets/editorial/court-1280.webp) U.S. Supreme Court, Washington, D.C. Photo: Carol M. Highsmith. Illustrative image. [Original source](https://commons.wikimedia.org/wiki/File:U.S._Supreme_Court_building-m.jpg) · [Public domain](https://creativecommons.org/publicdomain/mark/1.0/) · [Source & licence](https://reinouttebrake.com/data#image-credits) Four subscribers have alleged that leading AI companies coordinated to slow improvements to their paid products. Their September 2026 lawsuit challenges the alleged agreement between competitors; it is not a court finding that [AI safety](https://reinouttebrake.com/stories/anthropic-blocked-the-prompt-but-did-its-ai-safety-system-prevent-the-harm) cooperation is unlawful. On September 18, four customers filed an antitrust lawsuit against Anthropic, OpenAI, Google and SpaceXAI. They want to represent a nationwide class of paying subscribers. Their argument is surprisingly simple. Competition forces every AI company to make its models better. If one company slows down, it risks losing customers, talent and technological leadership. But if competitors agree to slow down together, customers may keep paying the same while receiving less improvement than competition would otherwise have produced. And that, the plaintiffs argue, can become an antitrust problem. The timing is not accidental. On September 12, Anthropic CEO Dario Amodei published “We Must Pace the Frontier”, arguing that AI capabilities are advancing faster than safety measures can keep up. But he went further. He said AI companies should voluntarily work together on [common safety standards](https://reinouttebrake.com/stories/agi-is-here-says-nvidia-nobody-is-ready-says-openais-chief-scientist) and limits on unchecked AI progress. He even acknowledged the antitrust issue, writing that some discussions would benefit from government mediation or a narrow antitrust waiver. Sam Altman, Elon Musk and Demis Hassabis then publicly backed the direction. The [lawsuit uses those public statements](https://news.bloomberglaw.com/ip-law/openai-anthropic-google-spacexai-hit-with-antitrust-lawsuit) as evidence that competitors agreed to reduce the pace at which their products improve. That is an allegation, not a finding: no court has established that an unlawful agreement existed. But the problem it exposes is real. Slow down alone, and competition punishes you. Slow down together, and competition law may punish you. Which means AI safety is no longer only an engineering problem. It is becoming a market-design problem. Should AI labs be allowed to slow down together: yes or no? ## Sources and context - [pacethefrontierlawsuit.com: source discussed in the editorial review](https://www.pacethefrontierlawsuit.com/) (Company or research source) - [darioamodei.com: source discussed in the editorial review](https://darioamodei.com/post/we-must-pace-the-frontier) (Company or research source) Filed under [AI](https://reinouttebrake.com/category/ai) ## Tags - [AI](https://reinouttebrake.com/tag/ai) - [ChatGPT](https://reinouttebrake.com/tag/chatgpt) - [Claude](https://reinouttebrake.com/tag/claude) - [gemini](https://reinouttebrake.com/tag/gemini) - [grok](https://reinouttebrake.com/tag/grok) Written by [Reinout te Brake](https://reinouttebrake.com/about) Gaming · AI · Tech · Entertainment · Dance Entrepreneur, consultant, investor and board member with 20+ years in tech and gaming, spanning business strategy, development, publishing, growth and investment. [Share on WhatsApp](https://wa.me/?text=AI%20Safety%20Antitrust%20Lawsuit%3A%20Can%20Rival%20Labs%20Slow%20Down%20Together%3F%0Ahttps%3A%2F%2Freinouttebrake.com%2Fstories%2Fai-was-supposed-to-slow-down-now-theyre-being-sued-for-it)[Share on LinkedIn](https://www.linkedin.com/sharing/share-offsite/?url=https%3A%2F%2Freinouttebrake.com%2Fstories%2Fai-was-supposed-to-slow-down-now-theyre-being-sued-for-it)[Share by email](mailto:?subject=AI%20Safety%20Antitrust%20Lawsuit%3A%20Can%20Rival%20Labs%20Slow%20Down%20Together%3F&body=https%3A%2F%2Freinouttebrake.com%2Fstories%2Fai-was-supposed-to-slow-down-now-theyre-being-sued-for-it) Cite this article Reinout te Brake. AI Safety Antitrust Lawsuit: Can Rival Labs Slow Down Together? (20 Sept 2026). [https://reinouttebrake.com/stories/ai-was-supposed-to-slow-down-now-theyre-being-sued-for-it](https://reinouttebrake.com/stories/ai-was-supposed-to-slow-down-now-theyre-being-sued-for-it) [Article data (JSON)](https://reinouttebrake.com/data/articles/ai-was-supposed-to-slow-down-now-theyre-being-sued-for-it.json)[Full article text](https://reinouttebrake.com/stories/ai-was-supposed-to-slow-down-now-theyre-being-sued-for-it.txt)[Markdown](https://reinouttebrake.com/stories/ai-was-supposed-to-slow-down-now-theyre-being-sued-for-it.md) Next / Read ## Related reading Follow the subject further. ![Dario Amodei, seated on the right, at TechCrunch Disrupt 2023](https://reinouttebrake.com/assets/editorial/amodei-1280.webp) [Image: TechCrunch](https://reinouttebrake.com/data#image-credits) More on AI safety Anthropic 12 Sept 2026 ### [Anthropic AI Safety: Blocking Prompts vs Preventing Harm](https://reinouttebrake.com/stories/anthropic-blocked-the-prompt-but-did-its-ai-safety-system-prevent-the-harm) Anthropic’s accounts of AI misuse expose different outcomes: refusing a prompt, disrupting an attack and preventing damage. Why that distinction matters. [Read analysis](https://reinouttebrake.com/stories/anthropic-blocked-the-prompt-but-did-its-ai-safety-system-prevent-the-harm) ![Jensen Huang speaking at Computex Taipei in 2016](https://reinouttebrake.com/assets/editorial/jensen-1280.webp) [Image: NVIDIA Taiwan](https://reinouttebrake.com/data#image-credits) More on AI safety Nvidia 7 Sept 2026 ### [OpenAI’s Chief Scientist Warns That AI Safety Is Falling Behind](https://reinouttebrake.com/stories/agi-is-here-says-nvidia-nobody-is-ready-says-openais-chief-scientist) Jakub Pachocki’s warning raises a practical question: what evidence and safeguards should businesses require before trusting more capable AI agents? [Read analysis](https://reinouttebrake.com/stories/agi-is-here-says-nvidia-nobody-is-ready-says-openais-chief-scientist) [More in AI](https://reinouttebrake.com/category/ai)[Rules, Legal Claims and Accountability](https://reinouttebrake.com/topic/ai#policy-and-accountability) [Browse all articles](https://reinouttebrake.com/articles)[Find a specific article](https://reinouttebrake.com/search) [Sources, dates & corrections](https://reinouttebrake.com/data#editorial-policy)