Bessent Says U.S.-China AI Dialogue Will Advance at APEC Summit
The proposed dialogue would cover advanced-model testing, human oversight and an incident-alert channel. Bessent said it is expected to advance at the Nov. 18-19 summit in Shenzhen.

Treasury Secretary Scott Bessent said the United States and China should coordinate on advanced artificial intelligence risks, with a new superintelligence dialogue expected to advance at next week’s Asia-Pacific Economic Cooperation summit.
Bessent made the remarks Nov. 10 at a National Committee on U.S.-China Relations gala in New York. He described the proposed dialogue as a place for both sides to exchange methods for assessing and testing advanced AI models, as well as keeping people in control of those systems. The discussions would focus on what Bessent called superintelligence, or SI.
Bessent also pointed to a direct alert channel intended to let the two governments notify each other about serious AI incidents. The channel was reportedly agreed during a White House meeting in September 2026. He cited it as one part of oversight for risks shared by the two countries.
Bessent said he would join President Donald Trump in Shenzhen, where Trump is due to meet Chinese President Xi Jinping on the sidelines of the APEC leaders summit. The summit is scheduled for Nov. 18-19, and Bessent said it would move the superintelligence dialogue forward. It would be the third Trump-Xi meeting this year; their first two meetings took place in Beijing and Washington.
The proposed dialogue reflects an effort to address shared AI risks while the two countries compete for technological leadership. The administration’s approach to China aims to reduce risky dependencies while maintaining ties. TokenPost previously covered U.S.-China AI discussions ahead of Xi’s September visit to Washington.
The Shenzhen summit is the next scheduled step for the dialogue, alongside the leaders’ meeting on its sidelines.