China urges US to collaborate more in the field of AI

The call for US-China AI cooperation on safety coincides with reports of Chinese AI models narrowing the performance gap with their American counterparts. A Bloomberg Intelligence report notes that top Chinese models now trail US rivals by just 3 percent on benchmark scores, a significant reduction from 15 percent earlier in the year. This indicates substantial progress by Chinese developers like DeepSeek, which could lead to increased market share.

Ambassador Xie Feng's argument emphasizes that open source and technological openness foster security, while decoupling and isolation introduce risks. This position is contrasted by the US initiative to create a "Super Intelligence Force" task force, announced by President Trump, which aims to ensure US leadership in AI. Despite these differing approaches, major AI companies in the US have also signed a voluntary accord for stronger internal controls and reviews of AI models.

Experts, including Sean Stein of the US-China Business Council, suggest that closer US-China cooperation on AI safety could serve as a stabilizing force in bilateral relations. Both nations share a definite interest in preventing malicious actors from manipulating AI, providing a concrete area where meaningful dialogue and collaboration could occur, irrespective of broader geopolitical dynamics. This shared concern could facilitate practical engagement on specific risk prevention measures.