China’s World AI Cooperation Organisation has added nine members in the month since it was founded, taking it from 29 signatories to 38, according to the South China Morning Post.
It is growing at precisely the moment Washington is drafting a letter telling 35 countries they cannot belong to both camps.
WAICO was signed into existence in Shanghai on 16 July by 29 countries, a year after Premier Li Qiang first proposed it at the World AI Conference.
Xi Jinping used the accompanying summit to argue that no single country should monopolise the technology, calling AI development “a symphony of international cooperation” rather than “a solo performance by a single country”.
The doctrine underneath the organisation has a name, and Beijing has now committed it to paper at the United Nations.
China’s Position Paper on Global Cyber Governance in the Digital Intelligent Age, released on 23 July, is organised around “respecting digital sovereignty and promoting development for all”.
Countries “should respect each other’s digital sovereignty, and they all have the right to independently choose their paths of digital intelligent technology development”, the paper says, and it went to the first meeting of the UN Global Mechanism on information and communications technologies in international security, held in New York from 20 to 24 July.
The formulation is not new so much as newly load-bearing. China’s Global AI Governance Initiative, launched at the Belt and Road Forum in October 2023, already asked states to “respect other countries’ national sovereignty” when supplying AI products and services abroad.
What has changed is that the phrase now has an institution behind it and an American initiative to define itself against. The State Department had drafted a letter to the 35 signatories of its AI Opportunity Statement warning them off “duplicative initiatives”, with the line “to be part of everything is to be part of nothing”.
Two caveats travel with that draft and are easy to lose. It is undated and, as of last week, unsent, and it names neither China nor WAICO anywhere in the text, though the push to make allies choose leaves little doubt about the target.
Pax Silica, the grouping the letter is meant to protect, launched in December 2025 with seven founders: the United States, the United Kingdom, Japan, South Korea, Singapore, Australia, and Israel. Its declaration now carries 25 signatories, while the separate AI Opportunity Statement was signed by 35 countries in June.
The arithmetic is what makes this awkward for Washington. WAICO’s 38 members and Pax Silica’s 25 are not mutually exclusive lists, and that overlap is the thing the draft letter exists to end.
Chinese commentary has been happy to name the strategy. “Washington wants a divided AI world by drawing a digital curtain to split the global market in two,” Li Yong of the China Society for WTO Studies told Global Times last weekend.
Beijing’s wider case rests less on treaties than on downloads. Alibaba’s Qwen family passed three billion downloads in six months with more than 300,000 derivative models, against 418 million for Google’s Gemma and 227 million for Meta’s Llama across 2026, per Hugging Face.
Open weights are a form of diplomacy that export controls struggle to reach. China has also committed 5,000 AI training places for developing countries, according to state media, a figure that has not been independently confirmed.
The vocabulary is shared even where the politics are not. Europe has spent two years making its own case for digital sovereignty, by which it mostly means control over its infrastructure and data rather than a right to be left alone by other governments.
There is also, in theory, a neutral venue. The UN established an independent scientific panel on AI and a global dialogue on AI governance in August 2025, and held the first dialogue in Geneva in July, while WAICO set up its secretariat in Shanghai and the American letter still has not been sent.