29 Countries Just Signed a China-Led Global AI Treaty. Here's What It Does
A new intergovernmental organisation called WAICO launched in Shanghai on 16 July 2026, with nearly three dozen countries signing on. The big question: will it actually change how AI is governed, or is it a talking shop?
Key points
- 29 countries signed the agreement creating WAICO, the World Artificial Intelligence Cooperation Organization, in Shanghai on 16 July 2026.
- The organisation is headquartered in Shanghai and operates as a China-led intergovernmental body, meaning it is run by governments, not a private group.
- Xi Jinping first proposed the organisation at an Asia-Pacific Economic Cooperation summit in October 2025.
- UN Secretary-General António Guterres attended the signing ceremony.
- Analysts disagree on whether WAICO will produce binding rules or remain a forum for discussion.
Twenty-nine governments gathered in Shanghai on 16 July 2026 and signed the founding agreement for WAICO, the World Artificial Intelligence Cooperation Organization, a new body designed to coordinate global policy on artificial intelligence. The ceremony took place the day before the annual World Artificial Intelligence Conference, the major yearly event where AI companies and governments showcase their latest work.
China's Foreign Minister Wang Yi signed on behalf of Beijing. UN Secretary-General António Guterres was present, lending the event a degree of international weight.
Who is actually in it?
Twenty-nine countries joined at launch in July 2026. Researchers who mapped WAICO's early membership describe its design as open to any sovereign state, with a particular focus on the Global South, the group of lower-income countries across Africa, Asia and Latin America that are often left out of decisions made in Silicon Valley or Brussels.
China first floated the idea in July 2025, and Xi Jinping repeated the call at the 33rd Asia-Pacific Economic Cooperation meeting that October.
What will it actually do?
That is less clear. Elizabeth Gibney, writing in the journal Nature in December 2025, speculated that WAICO is unlikely to govern the AI industry in an enforceable way. In other words, no fines, no binding bans, no teeth.
Arindrajit Basu of the Carnegie Endowment for International Peace suggested China may use the body to expand its network of AI cooperation centres abroad and to shape how AI gets discussed at the United Nations.
Digital policy expert Heru Sutadi, speaking about Indonesia's membership, put the challenge bluntly. The computing power, the specialist chips called GPUs, the cloud infrastructure, and the training data that AI systems need are all concentrated in a handful of large technology companies. Countries in the Global South risk becoming permanent consumers of AI products built elsewhere, he argued, without ever developing their own capabilities. He called on WAICO to actively defend those countries' interests rather than simply host conferences.
What does this mean for ordinary people?
For most people, nothing changes tomorrow. WAICO has no power yet to regulate apps, set safety standards, or force companies to behave differently. What it does create is a formal table where governments from outside Europe and North America can argue for a say in how AI develops globally.
Whether that table produces real rules, or simply produces more meetings, will become clearer over the coming months as WAICO holds its first working sessions from its Shanghai headquarters.
