US and China unlikely to curb AI development despite industry calls for slowdown

Industry leaders have urged a worldwide slowdown of artificial intelligence (AI) development, but the United States and China—responsible for most frontier AI labs—have shown no intention of curbing progress. The two countries are set to meet in Washington, D.C., for a third summit this year, with expectations limited to modest cooperation on safety measures.

Alvin Wang Graylin, a senior fellow at the Asia Society Policy Institute, said the United States and China host nearly all advanced AI labs and models, making any safety regime that covers only one side insufficient. He described agreed guardrails as a “precondition” for safe AI development.

U.S. President Donald Trump dismissed concerns about AI risks, calling them a “conspiracy” and arguing that guardrails are unnecessary under his administration. He told the United Nations General Assembly that Washington would oppose any “globalist scheme” to control AI, emphasizing the need for the United States to stay ahead of China in the technology.

China has issued a third version of its AI security framework but also rejected calls for a slowdown, labeling fears of existential threats as “fearmongering.” Chinese officials have framed AI risks in terms of cybersecurity and political stability. In an essay, State Security Minister Chen Yixin warned that hostile forces could use AI to spread false political rumors and incite antagonistic sentiments.

Experts note that mutual mistrust and differing views of the primary risks hinder cooperation. Sun Chenghao of Tsinghua University said the biggest obstacle is separating genuine safety collaboration from broader technology competition, adding that Beijing will question any safety talks linked to export controls or efforts to slow Chinese development.

So far, the only publicly discussed steps are a bilateral dialogue and a notification system for national-security threats. Analysts say further agreement would be needed on which AI capabilities should be off-limits to avert the greatest dangers.

Within the AI sector, opinions diverge on regulation. Some leaders, such as Anthropic CEO Dario Amodei and OpenAI CEO Sam Altman, have called for a slowdown and tighter oversight, while others, including Nvidia CEO Jensen Huang and Meta CEO Mark Zuckerberg, argue that external oversight could hinder the technology’s benefits.

Calls for an “indefinite international moratorium” on frontier AI have been made by academics like David Krueger, but many experts stress the need for coordinated safeguards rather than a full pause. Otto Barten of the Existential Risk Observatory warned that AI should never reach a level capable of recursive self-improvement, urging the United States and China to set clear red lines.

Both sides could build trust by separating safety policies from competitive strategies. Sun suggested the United States clarify its safety agenda, while China could increase transparency about its safety practices and incident-reporting mechanisms. Henry Gao, a trade-law specialist, noted that the United States focuses on catastrophic loss of control, whereas China views AI risks through a geopolitical lens, fearing a decisive U.S. advantage.

Some commentators argue that dialogue should concentrate on mitigating existing risks rather than halting development, emphasizing the technology’s potential to improve lives. Jiang Tianjiao of Fudan University called the upcoming talks a “positive signal,” expressing hope that the two leading AI nations could set a global example by managing risks together.