The channel through which the United States and China agreed in September to warn each other of serious incidents involving artificial intelligence has not yet begun to operate, according to several people familiar with the discussions who were cited in a Reuters report carried by the Singapore news site AsiaOne on 17 November 2026. Three people briefed on preparations told the same report that the two sides had planned further AI talks around the leaders' meeting in Shenzhen on Thursday 19 November. Neither government has announced a date for those talks in the documents this site read.1
What was agreed in September
The White House fact sheet of 25 September 2026, issued during President Xi Jinping's state visit to Washington, says the two countries set up a dialogue on the risks and benefits of the technology and agreed "to establish a bilateral communication channel for SI incidents". It adds: "The next exchange will occur by November 2026." SI stands for "super intelligence", the term the fact sheet uses in place of artificial intelligence.2
China's account, which a paper by Zhou Jinglun published by Fudan University's development institute attributes to a dispatch of 25 September by the state news agency Xinhua and quotes as item seven of an eight-point list, lists the same two items in different words:6
双方同意建立中美人工智能对话,交流人工智能相关风险和惠益。下次对话将于今年11月举行。双方同意建立人工智能事件的沟通渠道。
The two sides agreed to establish a China-US artificial intelligence dialogue to exchange views on the risks and benefits of artificial intelligence. The next dialogue will be held in November this year. The two sides agreed to establish a communication channel for artificial intelligence incidents.6
The passages this site read do not say which agencies run the channel, what counts as an incident or how quickly a warning must be sent.2 6
The two sides do not use the same name for it
The White House says the two leaders "agreed to use the term 'super intelligence' rather than 'artificial intelligence'"; China's published account keeps the words artificial intelligence. Asked on 26 September about the American wording, a Chinese Foreign Ministry spokesperson replied:2 4 6
关于人工智能表述问题,中方重视美方的立场,也尊重美方的提法。
On the question of how artificial intelligence is described, China attaches importance to the US position and also respects the US wording.4
Two days later, on 28 September, Xinhua reported the ministry's spokesman Guo Jiakun as saying that China holds that the United Nations should be the main channel for governing the technology, and that it is also willing to keep in contact with the United States through the government-to-government dialogue.5
What Washington says the dialogue will do
The US Treasury Secretary, Scott Bessent, described the work in remarks prepared for a speech in New York on 11 November 2026. He said the two governments expect to "exchange best practices for evaluating and testing advanced models before their release", and that they are "establishing a direct channel to ensure that either nation can promptly alert the other to serious incidents involving its SI models". He also said the two countries should "harbor no illusions" about their differences, and that he would join President Trump in Shenzhen.1 3
The Reuters report says the alert mechanism was discussed by Mr Bessent and China's Vice Premier He Lifeng in September and is meant to stop incidents or misunderstandings from growing into security crises. The people it cites did not say what is delaying it, and said contact so far has relied largely on direct exchanges between senior officials. China's Commerce Ministry and the US Treasury did not immediately answer the report's requests for comment.1
What a Chinese researcher expects
Zhou Jinglun, in a paper published by the Fudan Development Institute on 8 October 2026, sorts possible subjects into those where co-operation is likely and those where it is not. His view is that a treaty-style arrangement is not yet possible, and that the likeliest results between 2026 and 2028 are narrower ones, such as an emergency notification channel and technical alignment of model testing and incident reporting. That is one scholar's assessment, not a government position.6
No Board of Trade announcement is expected on Thursday
A senior US administration official told the same report that the two countries will not announce any further progress on the Board of Trade, the body considering lower tariffs on $30 billion of goods in each direction.1 3 Analysts quoted in the report said the prospect of tighter US chip export controls is likely to overshadow the meeting, though one said neither side wants it to sour the summit. China has named AI and the digital economy as priorities for the Asia-Pacific Economic Cooperation forum in Shenzhen, the report says.1
The White House text says the next AI exchange "will occur by November 2026", China's says it will be held in November, and the leaders are due to meet on 19 November. Whether the dialogue convenes in Shenzhen, and whether either side then says the incident channel is working, is not yet known.1 2 6