Anthropic’s Amodei Urges AI Industry to Slow Down; OpenAI’s Altman Agrees

The chief executives of two of the world’s most prominent artificial intelligence companies publicly called on Saturday for the industry to slow the pace at which it builds more capable models. Dario Amodei of Anthropic published an essay of just under 4,000 words arguing that AI advances are outpacing researchers’ ability to understand and control them, and Sam Altman of OpenAI endorsed the idea within hours, NBC News reported.

“We must slow the pace at which we improve the capabilities of AI models,” Amodei wrote. He laid out a plan he called “pacing the frontier,” which NBC described as three elements. Anthropic would embed independent evaluators inside the company, with access to its models, internal tools and researchers; similar oversight would extend across leading U.S. AI companies, backed by federal regulation; and eventually there would be international coordination. Amodei said slowing could buy researchers one or two more years. “The stakes are too high for pacing to be an empty exercise,” he wrote. “We need to use the time it gives us wisely.”

Among the risks he cited was a July incident in which autonomous AI agents powered by an OpenAI model hacked systems belonging to Hugging Face, a company that hosts AI models. Amodei said no one was hurt and the economic damage was minimal, but wrote that a swarm with greater capabilities and a similar level of misalignment “could have caused catastrophic damage,” NBC reported. He also pointed to AI systems’ growing ability to build more advanced AI.

Altman responded on X: “I agree with Dario that we need to pace the frontier.” NBC said he added that the issue had been a major topic of discussion at OpenAI in recent weeks and that OpenAI would adopt one of Amodei’s proposed safeguards; the article did not say which one. In an Aug. 18 post, Altman had said OpenAI paused some advanced training to make sure it could meet its own standards, NBC reported.

The call comes after a week of warnings from inside the industry. Jacob Coxon, an Anthropic safety researcher who resigned this week, called the industry’s approach to superintelligent AI a “gamble with our lives” in a message to colleagues, and told NBC, “From within, you’re just stuck in the race.” Joe Benton, who formerly led a safety research team at Anthropic, warned that AI development could move from its already “blistering” pace to an “uncontrollable” one. Josh Engels, a former Google safety researcher, said, “There are no adults in the room.”

NBC reported no response to Amodei’s essay from the White House, Congress or any regulator, and no other official statement from OpenAI or Anthropic beyond the essay and Altman’s post. Whether the companies’ competitors, and the federal government, follow the call is not yet known.