In a highly-anticipated talks held between the US Treasury Secretary Scott Bessent and Chinese Vice Premier He Lifeng on Sunday, AI safety proposals took center stage.
In the meeting, Bessent proposed a new AI safety notification mechanism ahead of the Trump-Xi summit scheduled for September 24. The newly proposed mechanism will be considered by both leaders at their meeting happening this week.
According to Bessent, discussions between the two nations centered on creating a new framework for artificial intelligence cooperation, highlighting national security priorities and a mutual alert mechanism for critical AI-related incidents.
“We think that just like with any cross-border activity, that moving from opaque to more transparency between the number one and the number two AI powers in the world is very important,” he added.
The proposal was put forward by the US side, however China’s response to this mechanism remained unclear. Even the readout published by the Chinese state news agency Xinhua did not mention this mechanism. The focus remained on in-depth discussions between officials over trade and key economic issues.
According to the experts, during the Trump-Xi summit, AI is expected to be high on the agenda. This meeting comes at a crucial time when researchers are sounding alarm over existential threats posed by artificial intelligence. In response tech industry leaders including Dario Amodei, Sam Altman and Elon Musk have called for slowing down AI development.
However, Trump dismissed such AI fears as a hoax and Chinese authorities also called them an act of fearmongering.
While trust remains low and cooperation is limited due to Beijing’s concerns over containment, both sides agreed to continue dialogue.Future sessions are expected to address sensitive issues, including the weaponization of AI, AI safety principles, critical infrastructure protection, and cyber-attack prevention.
US Representative Ro Khanna emphasized that Trump and Xi need to reach an international agreement including AI safety, specifically banning recursive self-improving AI and establishing safeguards against the misuse of AI for biological or nuclear weapons.
