one year on
China proposes World AI Cooperation Organization, unveils global governance plan at Shanghai conference
Premier Li Qiang calls for a multilateral AI governance framework as the US pursues a dominant unilateral approach, positioning China as a convener of Global South nations.
SHANGHAI — Chinese Premier Li Qiang today proposed the creation of a World Artificial Intelligence Cooperation Organization, with headquarters tentatively planned for Shanghai, alongside a 13-point Global AI Governance Action Plan, at the opening of the 2025 World Artificial Intelligence Conference (WAIC). The announcement frames a direct challenge to the US approach to AI governance, with Li warning that global AI oversight is fragmented and must not be monopolized by a few countries or companies.
“Overall, global AI governance is still fragmented,” Li said. “Key resources and capabilities are concentrated in a few countries and a few enterprises.” The proposal arrives days after the US released its own AI Action Plan emphasizing American dominance. Li’s speech stressed universal access, collaborative innovation, and collective governance, explicitly courting Global South nations with pledges to share development experience and technology.
The conference drew more than 1,000 domestic and international guests from more than 40 countries, including Turing Award winner and Nobel laureate Geoffrey Hinton, making his first visit to China. Hinton discussed the risks of being overtaken by intelligent systems and was deeply involved in the “Shanghai Consensus” on AI safety, which calls for safety assurances from frontier developers and verifiable global red lines.
The record
Called on the US and China to collaborate to keep the world stable, peaceful and under human control.
view the original post →Discussed the risks of being overtaken by intelligent systems and joined the "Shanghai Consensus," which calls for safety assurances from frontier developers, verifiable global red lines, and investment in safe development.
view the original post →One year later — open only if you can handle spoilers
One year later, on 16 July 2026, China formally launched WAICO in Shanghai, with 29 nations — including Russia, Pakistan, Indonesia, Kazakhstan and Laos — signing on as founding members; UN Secretary-General António Guterres attended the ceremony and Foreign Minister Wang Yi signed for China. The United States and most Western governments stayed out, leaving the body weighted toward the Global South, while Beijing pledged to train 5,000 practitioners, release open-source models and fund AI infrastructure for member states over five years. Rather than converging, the rival US and Chinese approaches to AI governance hardened into competing blocs.
The Weekly Replay · free by email
This week, one year ago — every Sunday.
One email each Sunday: the week's replayed AI news, with the one-year-later annotations included. Written like it's breaking — dated like it isn't.
Free · double opt-in · unsubscribe anytime · privacy