AI safety concerns reach a peak in the US after a X contribution by a former Anthropic researcher went viral claiming the technology could kill all of humanity by the end of the decade. In response, the CEOs of OpenAI and Anthropic have reiterated their calls for the US to coordinate with China to slow and speed up AI development.
Sam Altman, CEO of OpenAI, even tried to appeal to the egos of the leaders. tell Assets in an interview on Friday He believed that US President Donald Trump and Chinese President Xi Jinping could win the Nobel Peace Prize if the two leaders reached an agreement on AI safety.
But the idea of AI slowing appears to be falling on deaf ears with President Trump and President Xi, who will meet on September 24. Both rejected the idea.
Trump said in a Sept. 14 Truth Social: “The only checks or ‘guardrails’ the US needs is a STRONG AND SMART (high IQ!) PRESIDENT.” post. On the same day, Chinese Foreign Ministry spokesman Guo Jiakun said called The current discourse in the US is a “fear-mongering” that “will only hinder efforts for solid global AI governance that serves no one.”
In China, calls for a slowdown also appeared as an attempt by the US to maintain its lead, as it tried to do by restricting exports of advanced AI chips to China.
“Xi is unhappy with recent U.S. moves to curb Chinese advances in AI, robotics and drones,” said George Chen, partner and chair of digital practice at The Asia Group. “For Xi, AI is the new Internet – a once-in-a-lifetime opportunity to reshape the technological balance of power. China doesn’t need US approval to speed up or slow down its AI investments; Xi will pursue his own agenda.”
Trump echoed Xi in his Truth Social post, saying, “Whoever wins AI wins!”
Leaders are expected to begin discussing AI security when they meet, but any kind of agreement between countries is “still a long way off,” said Paul Triolo, head of global technology policy at consultancy DGA. Xi wants to have a “serious dialogue on the security of AI models in border areas,” Triolo said, but the two countries still have “a basic level of agreement on things like the role of government, [and] how and which models should be tested.”
Also under discussion are an agreement not to use AI as a weapon and a study of “principles to prevent misuse of AI models by non-state actors, such as attacks on global financial systems or critical infrastructure, which no country wants to see,” Chen said.
The letter from Anthropic CEO Dario Amodei was poorly received in China – and by Trump
Although OpenAI CEO Sam Altman tweeted Anthropic CEO Dario Amodei’s letter published on September 12th about coordinated mitigation and asking the US government to facilitate it sparked a particularly polarizing reaction. While some in the US praised it as a useful framework for mitigating the risks of AI, it was not well received in China.
In the letterAmodei calls on the US and China to agree on a “speed limit” for AI development. When discussing his letter in an interview with CBS Sunday Morning, Amodei compared the competition between the U.S. and China to the Cold War, when the U.S. and the Soviet Union fought to develop nuclear weapons. Amodei says Washington DC and Beijing should at least agree that neither country will use AI to develop biological weapons, and he reiterates his belief that the US should not sell advanced chips to China.
Brosi Babic, a professor at the University of Hong Kong, calls Amodei’s letter a “self-serving editorial” that “aptly comes at a time when the gap between Chinese and American high-end models is narrowing, as an attempt to hold on to a dwindling market lead.” In his opinion, the proposals to slow down AI are “so insidious and childishly worded” that they are unlikely to advance Xi’s agenda for the meeting with Trump.
Trump also denounced Amodei’s letter. He called Nvidia CEO Jensen Huang when Huang happened to be speaking on stage. Huang put Trump on speakerphone and Trump in front of the crowd said“Whatever Dario said this weekend, it won’t stop our progress.” He also called the backlash against data centers and fears that AI will “take over” a “hoax.”
A social media post by Shengyu Liu, an engineer at DeepSeek, gains traction because he compares Anthropic achieving “advanced artificial intelligence” to “Hitler achieving atomic bomb technology before the Allies.”
Liu also highlights another key difference between the U.S. and China’s approach to AI technology: The U.S. industry generally prefers closed-weight models from companies like Anthropic and OpenAI, while China has focused on releasing lower-cost, open-weight models like those from DeepSeek.
“I still believe that border information should be made available to everyone in an open and cost-effective manner,” Liu said. “I don’t trust Anthropic or OpenAI to do this.”
The AI dialogue in China is dramatically different than that in the US
Outside of politics, the current uproar in the US over AI safety and “saving humanity” – a phrase appropriated by the tech industry – has not caught on in China. In fact, the Chinese public also generally views the idea of a slowdown as a U.S. attempt to get ahead and has a growing distrust of the U.S. tech industry.
“In particular, younger generations are adopting more pro-government views and are encouraged to be proud of the ‘new generation of Chinese’, with the ‘China rising, US falling’ narrative gaining traction,” Chen said.
Most Chinese believe it is the government’s responsibility to manage AI security risks, Chen said. For ordinary citizens and business people, it’s less about regulation and more about the practical benefits of AI, including how they can use it to improve their daily lives or generate income.
“The vast majority of average Chinese citizens are very positive about technology in general and AI in particular,” Triolo said. “They’ve seen significant improvements in the quality of life in China through technology and are very willing to embrace AI and other technologies. Some of the worst issues in AI, such as existential threats to humanity, aren’t really discussed much.”
However, those working in the Chinese AI industry are aware of the security risks and take them seriously. “The levels of discussion within this group are very sophisticated,” Triolo said. “Chinese AI security researchers are very concerned about cybersecurity and biosecurity risks, as well as things like loss of control.”
Some Chinese politicians have also called for more regulation, although not necessarily in partnership with the US. Chen Yixin, the head of China’s Ministry of State Security, this week called for greater government oversight of AI, which he said poses a direct threat to the Chinese Communist Party’s control. The New York Times Reports.
Given Trump and Xi’s apparent rejection of the idea of slowing AI, and the growing gap between nations’ attitudes toward AI, all eyes will be on next week’s Trump-Xi meeting to see what progress, if any, can be made to avoid the potentially catastrophic outcome that US labs are warning of.