US, China to hold AI talks in September, say sources

BEIJING – The US and China will hold talks over artificial intelligence in September, five people familiar with the matter told Reuters, as the rival superpowers grapple with how to mitigate the risks posed by each other’s increasingly powerful frontier models.

The AI talks, a significant outcome of the Trump-Xi summit in May, will likely take place before Chinese President Xi Jinping’s planned Sept 24 visit to the US, said the sources, who asked not to be named as the information is not public.

The dates, however, have not been finalised, they added.

As economic, political and military rivals, China and the US have grown increasingly anxious over each other’s AI capabilities amid an accelerating global AI race.

Beijing and Washington are weighing how to regulate advanced AI models that could strengthen military capabilities, enable cyberattacks on critical infrastructure and disrupt labour markets.

The September talks, reported by Reuters for the first time on July 21, would be led on the US side by Treasury Secretary Scott Bessent, four of the sources said.

The identities of other participants from the US and China, as well as the agenda and location of the discussion, are still under deliberation as preparations are in the initial stage, three of the people said.

The White House, the US State Department and the US Treasury did not immediately respond to a request for comment.

China’s National Development and Reform Commission, cyberspace regulator, industry and foreign ministries, and the State Council did not immediately respond to faxed requests for comment.

Bessent, asked about the talks in an interview on Fox Business Network on July 21, did not directly confirm the meeting but cited concerns about intellectual property and said models need to be held to the same standards.

“We are finding watermarks of our US large language models on many of the Chinese models, and that that’s unacceptable,” he told the US television network.

“We’re going to be looking at that in the coming days or weeks.”

Talks will be first official AI dialogue under Trump

Former US president Joe Biden agreed with Xi that humans, not AI, should control decisions on the use of nuclear weapons following working-level talks in Geneva in 2024.

But the planned meetings would constitute the first official US-China dialogue on AI under President Donald Trump.

The U.S. has already limited exports to China of some of the most advanced AI chips, but has held off on adding China’s AI start-up DeepSeek, memory chipmaker CXMT and more than 100 other companies flagged as national security risks to a trade blacklist in an effort to avoid escalating tensions with Beijing, Reuters has reported.

Asked about possible trade actions, US Trade Representative Jamieson Greer said a number of US government agencies were looking at the issue. He cited concerns over forced technology transfer and intellectual property issues as well as energy subsidies but gave no other details.

“We’re taking a very close look at how China is propagating its AI development to make sure that our companies compete... on a level playing field,” Greer told CNBC in an interview.

The Chinese authorities, meanwhile, are mulling over curbs on overseas access to the most powerful Chinese-developed AI models, amid concerns about the hacking potential of Anthropic’s Mythos, which is currently not available to the general public.

In May, Bessent said that future talks would be “aimed at halting proliferation of powerful AI models... to non-state actors” and that meetings could start “within the next four to eight weeks”.

Following Trump’s visit, China’s Foreign Ministry confirmed that the two nations had agreed to establish intergovernmental AI talks, but Beijing has not commented publicly since.

China looking for engagement, not political talks

Beijing’s priority list for the discussions includes Mythos, how the US will control the release of future advanced frontier models, and whether Washington will impose restrictions on Chinese open-weight models, analysts said.

The Trump administration is in talks with US AI companies about creating voluntary safety standards for the release of new AI models, but it has not yet announced any regulatory measures.

The Chinese authorities will continue internal deliberations on AI issues over the next month before deciding which officials from which ministries will participate in the AI dialogue, one of the sources said.

China is looking for real engagement and technical discussions on AI issues from the US rather than political talks, the source added.

“This is the first meeting; they’re unlikely to solve any major problems. I think they will try to agree to some basic terms, like what defines a frontier AI model,” said Paul Triolo, a partner at DGA-Albright Stonebridge Group. REUTERS