On July 16, representatives of twenty-nine countries signed the founding agreement for the World Artificial Intelligence Cooperation Organization in Shanghai. The next morning, Xi Jinping delivered his first in-person keynote at the World Artificial Intelligence Conference, the annual event China has staged since 2018. United Nations Secretary-General António Guterres attended the signing. Chinese Foreign Minister Wang Yi signed on behalf of Beijing. The plan was a year old. Premier Li Qiang had unveiled it at the same venue in July 2025 and spent twelve months securing commitments. The organization will be headquartered in Shanghai, the first intergovernmental body devoted specifically to artificial intelligence with a permanent physical seat.
The founding members are Algeria, Belarus, Brazil, Cambodia, Cameroon, China, Congo, Cuba, Ethiopia, Indonesia, Kazakhstan, Kenya, Kyrgyzstan, Laos, Lesotho, Malaysia, Mozambique, Myanmar, Nicaragua, Oman, Pakistan, Russia, Senegal, Serbia, South Africa, Tajikistan, Uzbekistan, Venezuela, and Zambia. Together they represent roughly three billion people. None is a G7 member. None is a Five Eyes partner. The European Union is absent. India is absent. The organization that will write AI governance standards for a third of the world's population does not include the country that builds the chips, the country that designs the models, or any of the countries whose regulatory frameworks currently govern the global technology industry.
The Empty Charter
WAICO's founding agreement specifies almost nothing. There are no published voting rules. There is no disclosed budget. There are no accession requirements, enforcement powers, or compliance mechanisms. The agreement does not explain how standards will be set, who will fund the secretariat, or what relationship WAICO will have with existing bodies like the OECD, the ITU, or the United Nations' own Global Digital Compact. Researchers at the arxiv mapped WAICO against fifteen existing AI governance instruments and found it occupies a position no other body holds: open membership with no ideological gatekeeping and a development-first agenda. Every other multilateral AI framework either restricts membership to like-minded democracies or anchors its mandate in human rights. WAICO does neither.
The Carnegie Endowment's Arindrajit Basu assessed that China "hopes to receive buy-in for its state-centric governance vision of technology from the Global South." France warned that WAICO could normalize authoritarian AI governance models and undermine the G7's Hiroshima Process. The organization that has no rules is already being criticized for the rules it might write.
The Offering
What WAICO does have is an offering. Xi announced five thousand AI training and seminar positions for developing countries over five years. China will deploy MAZU, its AI-powered meteorological early-warning system, across thirty nations within five years. International AI application cooperation centers will be established with ASEAN, the Arab League, the African Union, the Community of Latin American and Caribbean States, the Shanghai Cooperation Organization, and BRICS. The pledges are infrastructure, not regulation. Training, weather systems, and cooperation centers are the vocabulary of development assistance, not governance.
The conference that hosted the signing reinforced the point. WAIC 2026 drew over 1,100 companies, featured more than 300 global product debuts, and covered 100,000 square meters of exhibition space. Shanghai's AI industry generated 637 billion yuan in revenue in 2025, growing 39.5 percent year-over-year. The city is targeting one trillion yuan by 2030. Huawei debuted its Atlas 950 SuperPod, the largest AI supernode. More than two hundred humanoid robots occupied a dedicated zone. The procurement platform expects to unlock 20 billion yuan in real demand. WAICO launched inside a trade show, not a diplomatic summit. The venue tells you what the product is.
The Contrast
Six weeks earlier, on June 2, President Trump signed an executive order on AI that explicitly bars mandatory oversight. The order asks AI companies to voluntarily submit their most powerful models for government testing up to thirty days before public release. It states that nothing in the order "shall be construed to authorize the creation of a mandatory governmental licensing, preclearance, or permitting requirement for the development, publication, release, or distribution of new AI models." The voluntary framework includes classified benchmarks administered by the NSA and a government-managed pre-release review window, but participation is elective. The United States chose not to build an institution.
China built one. The contrast is not voluntary versus mandatory. WAICO has no binding power either. Both approaches are voluntary. The difference is architecture. The American approach assumes its companies will lead and the world will adopt their standards by default. Market dominance equals standard-setting. OpenAI, Anthropic, Google, and Nvidia set de facto standards through their products. The Chinese approach assumes the opposite: that the United States will not share, and that a parallel system must be built. WAICO sets standards through its charter, not its products. One side bets on market power. The other bets on institutional membership.
The historical record favors de facto standards. VHS beat Betamax. TCP/IP beat OSI. The dollar beat the bancor. But there is one category where de jure standards reliably win: industries where the product requires government permission to deploy. Telecommunications. Pharmaceuticals. Banking. Energy. AI is moving into that category. If WAICO's twenty-nine member states require WAICO compliance for domestic AI deployment, Chinese AI companies will hold a regulatory passport that American companies do not.
The Vacancy
For eighty years, the United States set global standards by building the institutions that housed them. ISO, ICANN, SWIFT, the Basel accords, the WTO's TRIPS agreement. American companies wrote the technical specifications. American diplomats wrote the governance frameworks. American lawyers wrote the compliance requirements. The architecture was American. Other countries joined because the alternative was exclusion from the system.
WAICO exists because the United States stopped building. The Hiroshima Process produced principles. The OECD updated its AI guidelines. The EU enacted the AI Act. But no Western-led body has created a standing intergovernmental organization with a physical headquarters, a membership roster, and an open door. WAICO filled the vacancy. Its authority is unproven. Its standards are unwritten. Its enforcement is nonexistent. But it has something no Western AI governance effort has produced: an address.
The question for the next twelve months is not whether WAICO can enforce standards but whether twenty-nine governments begin requiring WAICO-aligned compliance as a condition of AI deployment in their markets. If they do, the organization that currently has no rules will have the only thing more valuable than rules: a list of countries where your product cannot ship without its stamp. That is not governance. It is a passport. And for now, only one country's companies are standing in line.