Anthropic's Amodei says China presents 'toughest dilemma' for his proposed AI slowdown

Anthropic CEO Dario Amodei said Sunday that the "toughest dilemma" about his proposal to slow the advancement of artificial intelligence is what happens if adversarial nations, namely China, choose not to do the same. "The more long-term thing would be working together to put a speed limit on the rate of of AI progress," Amodei told CBS News' "Sunday Morning." "I think that's going to be very difficult because the incentives to pull ahead and the military advantage that you get from that are so large. And honestly, I don't know if it's possible, but we should we should try." Amodei published an essay on Saturday urging AI companies to slow how quickly they improve their most advanced models, a move that followed stark warnings from industry researchers last week about the technology's growing potential to cause catastrophic harm. An Anthropic researcher, Jacob Coxon, announced he quit his job out of concern that the company, and its chief rival, OpenAI, are "gambling with our lives." Amodei's essay landed during a pivotal time for the AI industry, which is facing public backlash over its data centers and mounting calls for regulation out of Washington. Anthropic and OpenAI are also gearing up for potentially historic IPOs, though neither company has set an official date for a debut. OpenAI CEO Sam Altman said Saturday that the company will likely not go public this year, in part because of concerns around safety. "Right now would be an ill-advised moment to go public," Altman said in an interview with Fortune. Altman's relationship with Amodei, who defected from OpenAI to start Anthropic, has been strained in recent years. But Altman expressed support for Amodei's essay in a post on X on Saturday, writing that he agrees on the need to "pace the frontier." Several other industry leaders, including Demis Hassabis, the chair of Google DeepMind, and Elon Musk, CEO of both Tesla and SpaceX, also welcomed Amodei's proposal. "Dario's essay points towards the right path forward," Hassabis wrote in a post on X on Saturday. "The details need working through, but the direction is correct for meeting this critical moment." In the essay, Amodei proposed a three-step plan aimed at slowing the pace of development without "sacrificing commercial advantage or the United States' lead in AI." Amodei said Anthropic has "unilaterally" committed to the first step of the plan, which grants third-party evaluators employee-level access to the company to verify safety practices and report incidents. The second step encourages leading AI companies within democratic countries to coordinate and establish common safety standards, and the third calls for coordination between democratic governments and authoritarian governments. "Dario is right that there should be some oversight," Musk wrote in a post on X early Sunday morning. "Peer review of AI by competitors is the right way to start this off." Musk, who has for years warned about the potential risks of AI, was quick to express support for Amodei on Saturday. He had been a major critic of Anthropic until his rocket company SpaceX, which acquired his AI startup xAI earlier this year, inked a major compute 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 highly competent and cared a great deal about doing the right thing," Musk wrote following the deal. "No one set off my evil detector." Alexandr Wang, Meta's AI chief, who joined the company as part of a $14 billion investment in his startup, Scale AI last year, has yet to address Amodei's essay directly. He shared a post on X on Saturday where he said Meta Superintelligence Labs, the company's AI research unit, is "rapidly scaling up the share of our efforts that goes into alignment as our models become more powerful." In the AI industry, alignment refers to the work by AI developers to ensure that the system behaves in accordance with human values and intentions. "We do believe alignment can be the gating factor for scaling as we get closer to the frontier," Wang said. David Sacks, a vocal critic of Anthropic who previously served as President Donald Trump's AI and crypto czar, also weighed in on Amodei's essay on Saturday. He wrote in a post on X that if OpenAI and Anthropic truly believe their models are "scary enough" to warrant a slowdown, he supports their "decision to be responsible." Even so, he said both companies need to "stop pretending you need anyone else's permission," and encouraged them to acknowledge that their motivations are not purely altruistic. He said they face "massive product-liability exposure" if their products were to enable something like a damaging cyberattack, which means "it is simply good business for OpenAI and Anthropic to trade some raw power for reliability and predictability." "So go ahead and pace the frontier," Sacks wrote. "You are the ones setting it. The easiest way not to build superintelligence is for you to agree not to build it. Demanding your preferred regulatory framework as the price of that will look like blackmail of the public and the political system. So just do it." WATCH: Seems like Anthropic will beat OpenAI to IPO, says FirstMark’s Rick Heitzmann