The CEO of Anthropic, Dario Amodei, suggested on Sunday that the United States and China should hold talks about artificial intelligence (AI) to lower the risks of “potential destruction.”
In a lengthy essay published Sept. 12, Amodei said that simply investing in risk prevention is no longer enough when it comes to AI.
He reiterated his concerns in an interview on “CBS Sunday Morning,” but CBS correspondent Jo Ling Kent asked him whether slowing down was a good idea for U.S. companies “when China will not” slow down.
“This is the toughest dilemma,” Amodei replied. “This is the worst aspect of our situation. This is the thing that most puts us in a bind. It’s not about commercial competition; this is the geopolitical situation which is the hardest to deal with.”
“We should talk with the Chinese and see if we can work together to put a speed limit on the rate of this technology,” Amodei said, drawing a parallel with the nuclear weapons disarmament talks between the United States and the Soviet Union during the Cold War.
The Chinese Ministry of Foreign Affairs’ spokesman, Guo Jiakun, made his own intervention into the debate during a regular news conference in Beijing on Monday.
China Cites ‘Fearmongering’
“Fearmongering, confrontation and vicious competition will only disrupt the process of global AI governance which serves no one’s interest,” Guo said.
President Donald Trump and Chinese leader Xi Jinping are due to meet at the White House on Sept. 24, and the governance of AI is one of a number of topics that will be on the agenda.
On Sunday, Trump was asked about the debate over AI during his tour of Ireland.
“We’re leading China in AI,” Trump told reporters at a golf course on Sept. 13. “We’re the most sophisticated country in the world, and frankly I want to keep it that way because whoever wins AI, wins.”
Trump said guardrails are an option.
OpenAI CEO Sam Altman (L), with U.S. President Donald Trump (C) and DeepMind CEO Demis Hassabis (R) at a working lunch during the G7 Summit at Evian-les-Bains, France, on June 17, 2026. Anna Moneymaker/Getty Images
“We can do this and that,” Trump said. “But I think you have a lot of very negative forces that are bringing it up that shouldn’t be bringing it up and they’re bringing up things that won’t happen.”
AI stocks globally slipped on Monday, and the Nasdaq Composite and S&P 500 indexes opened lower.
The ongoing debate was reignited on Sept. 9, when Anthropic’s Alignment Science lead, Evan Hubinger, posted on X in response to a tweet by AI researcher Jacob Coxon, who wrote, “The people building AI earnestly believe that it could kill us all by the end of the decade.”
“Jacob is correct here—we really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade,” Hubinger wrote.
“I believe Anthropic is trying its best, but we do not yet have a plan to solve alignment for superintelligence and are not clearly on track to.”
Amodei then published his essay on Saturday, in which he emphasized one point.
“We must slow the pace at which we improve the capabilities of AI models,” he wrote. “Progress will still seem fast, and we must make wise use of the time we gain.”
But Amodei said that slowing down, or pacing, had to be limited by the lead U.S. companies have over authoritarian regimes like the Chinese Communist Party (CCP).
“If we slow down by more than this amount, then (unpaced) CCP-associated projects will pull ahead, creating significant national security risk,” Amodei said, adding that he agreed with Treasury Secretary Scott Bessent, who said last week that the danger of China winning the AI race posed an existential threat.
‘No Day After Tomorrow’ If China Wins
“Beating China—there is no day after tomorrow if China wins at this,” Bessent said on Sept. 9 at Breitbart News’s “State of the Economy” event in Washington.
“If they were to pull ahead of us on AI, then nothing else matters.”
Overnight, the CEO of one of Anthropic’s commercial rivals Open AI, Sam Altman, appeared to support Amodei’s call for a slow down.
“When we talk about ‘pacing’, we do not mean ‘stopping’,” he posted on X.
“Progress has been rapid and will continue to be. But it should be slower than it otherwise could be; interventions like safety cases and monitoring have significant costs.
“The world deserves confidence that American companies developing increasingly capable AI will act responsibly, especially as the trajectory of progress has steepened.”
The Associated Press contributed to this report.
English (US) · 
Turkish (TR) · 