Anthropic CEO Dario Amodei at the AI Impact Summit on Thursday, February 19, 2026 in New Delhi, India.
Ruhani Kaur | Bloomberg | Getty Images
human CEO Dario Amodei said Sunday that the “toughest dilemma” about his proposal to slow advances in artificial intelligence is what would happen if an adversary, namely China, chooses not to do the same.
“Longer term, I think we’ll work together to put a rate limit on how quickly AI advances,” Amodei said on CBS News’ “Sunday Morning.” “I think that’s a very difficult thing to do because the incentive to move forward and the military advantage that comes from it is so great. And to be honest, I don’t know if it’s possible, but we should try.”
Amodei published an essay on Saturday urging AI companies to slow down the pace at which they improve their cutting-edge models, after industry researchers issued stark warnings last week about the technology’s increasing potential for catastrophic damage. Human researcher Jacob Coxon believes that the company and its biggest rival OpenAIit’s a “gamble with your life.”
Amodei’s essay comes at a pivotal time for the AI industry, which is facing public backlash over data centers and growing calls for regulation from Washington. Anthropic and OpenAI are also gearing up for potentially historic IPOs, but neither company has set a formal debut date. OpenAI CEO Sam Altman said Saturday that the company likely won’t go public this year, due in part to safety concerns.
“This is an unwise time to go public,” Altman said in an interview with Fortune magazine.
Relations between Altman and Amodei, who left OpenAI to start Anthropic, have been strained in recent years. But Altman expressed support for Amodei’s essay in a post on X on Saturday, writing that he agreed with the need to “pace the frontier.”
Several other industry leaders, including Demis Hassabis, Chairman of google DeepMind and CEO Elon Musk tesla and space xalso welcomed Mr. Amodei’s proposal.
“Dario’s essay points to the right path forward,” Hassabis wrote in Saturday’s X. “Details need to be worked out, but we are on the right track to address this critical moment.”
In his paper, Amodei proposed a three-step plan aimed at slowing the pace of development without “sacrificing commercial advantage or America’s lead in AI.”
Amodei said Anthropic is working “unilaterally” on the first phase of the plan, giving third-party evaluators employee-level access to the company to examine safety practices and report incidents. The second step would encourage major AI companies within democracies to coordinate and establish common safety standards, and the third step would require coordination between democratic and authoritarian governments.
“Mr. Dalio is right that there must be some oversight,” Musk wrote in a post to X early Sunday morning. “Peer review of AI by competitors is the right way to start this.”
Musk, who has long warned about the potential risks of AI, was quick to voice his support for Amodei on Saturday. He was a major critic of Anthropic until SpaceX, the rocket company that acquired his AI startup xAI earlier this year, signed a major computing deal with the company in May. As part of the deal, Anthropic will pay SpaceX $1.25 billion per month through May 2029, according to filings.
“Everyone I met was extremely competent and extremely concerned about doing the right thing,” Musk wrote after the deal. “No one set off my evil detector.”
Alexander Wang meta The AI chief, who joined the company last year as part of a $14 billion investment in his startup Scale AI, has yet to directly address Amodei’s essay.
He shared a post on Saturday’s X, saying the company’s AI research arm, Meta Superintelligence Lab, is “rapidly increasing its share of coordination efforts as models become more powerful.” In the AI industry, coordination refers to the work done by AI developers to ensure that systems behave in accordance with human values and intentions.
“We believe that as we approach the frontier, coordination can be a determining factor in scale,” Wang said.
David Sachs, a vocal critic of Anthropic and former President Donald Trump’s AI and cryptocurrency czar, also weighed in on Amodei’s essay on Saturday. In a post on X, he wrote that if OpenAI and Anthropic truly believe their models are “sufficiently scary” to warrant slowing down, he supports their “responsible decisions.”
Still, he said both companies need to “stop pretending they need other people’s permission” and urged them to recognize that their motives are not purely altruistic. He said they would face “massive product liability risks” if their products enabled something like a harmful cyberattack, which meant “for OpenAI and Anthropic, trading some of their raw power for reliability and predictability is simply good business.”
“So move on to the frontier,” Sacks wrote. “You’re the ones setting it up. The easiest way not to build a superintelligence is to agree not to build one. Demanding the regulatory framework you want in return would seem like a threat to the public and the political system. So just do it.”
WATCH: Anthropic looks set to beat OpenAI to IPO, says FirstMark’s Rick Heitzmann
