The United States and China plan to establish a channel for rapid warnings about serious AI incidents, U.S. Treasury Secretary Scott Bessent said, as the governments begin talks on managing advanced-model risks.
Bessent said President Donald Trump directed the dialogue on superintelligence. Speaking Tuesday, Nov. 10, at a National Committee on U.S.-China Relations gala in New York, he said the countries expect to compare how they review advanced models before deployment and keep people responsible for oversight. The Treasury published his remarks.
The planned channel would let either government quickly alert the other to a serious incident involving its models, Bessent said. Bessent said the broader dialogue was designed to “guard against shared risks.” He also cautioned that cooperation would not mean their interests always align.
China’s Ministry of Commerce (MOFCOM) and Ministry of Science and Technology (MOST) have proposed adding 23 AI-related technologies to the export-control catalogue. In an official Q&A, the ministries said the proposed changes are intended to address risks that AI technology could escape control or spread improperly, while allowing normal international exchanges and cooperation. EconoTimes previously covered China’s proposed AI export-control changes.
Bessent also said he would join Trump in Shenzhen the following week.
