29 countries sign agreement founding China-led World AI Cooperation Organization

A CSIS paper reviews three institutional approaches to global AI governance that advanced within about six weeks: the second Pax Silica Summit hosted by the US State Department in June, the UN Global Dialogue on AI Governance in Geneva, and the signing on 16 July 2026 in Shanghai of an agreement creating the World Artificial Intelligence Cooperation Organization (WAICO) by representatives of 29 countries. The paper argues the United States is in an institutional race with China over who writes the rules, and links its analysis to a US-China summit planned for 24 September.

According to the paper, WAICO is a China-headquartered intergovernmental body operating outside the UN system, and no G7 economy or EU member joined. Its 29 signatories are mostly low- and middle-income Global South countries. The paper traces the path from China's 2023 Global AI Governance Initiative and a 2024 UN resolution on capacity-building to a 2025 proposal for a permanent body. Xi Jinping then announced 5,000 AI training opportunities over five years. By contrast, Pax Silica, begun in December 2025 with 11 signatories and at 24 by June, is built around supply chain security, not governance.

The paper compares WAICO with the Belt and Road Initiative and the Asian Infrastructure Investment Bank, suggesting possible rapid uptake followed by concern over dependency, and noting that an OECD member joining could prompt others to follow. Kazakhstan joined both WAICO and Pax Silica, and membership, the paper says, should not be read as endorsing China's domestic model.

It cautions that WAICO has no binding commitments or standards yet, and little is known about its secretariat, budget or decision rules. Its recommendations call for a durable national framework, mutually acceptable verification, and bilateral outcomes that carry into multilateral forums.