Xi Jinping walked onto the stage at the Shanghai Expo Center on Friday and, for the first time in the eight years the World Artificial Intelligence Conference has convened, delivered its keynote in person. The choreography was the message. A day earlier, 29 governments had signed an agreement establishing a new intergovernmental body headquartered in Shanghai — and Xi's speech served as its coming-out party, a bid to organize much of the world's AI governance around Beijing rather than Washington or Brussels.
The new organization is the World Artificial Intelligence Cooperation Organisation, or WAICO. Its founding members read like a roster of the Global South and China's strategic partners: Indonesia, Brazil, Malaysia, South Africa, Senegal, Pakistan, Russia, Kazakhstan, Belarus, Serbia, Cuba, Venezuela and Laos among them. The United States is not on the list. Neither is any member of the European Union's core. That absence is not incidental — it is the point.
What was announced
"AI development should not be a solo performance by a single country, but a symphony of international cooperation," Xi told the conference, according to reporting from Al Jazeera, France 24 and Business Standard. He called the new body "an important milestone in the history of AI development."
Xi paired the pledge of cooperation with a pointed jab at the export-control regime the US has built around advanced chips. "We should jointly oppose overstretching the national security concept in the field of AI, or placing one country's security over that of others," he said — language that landed squarely on the Biden- and Trump-era restrictions that have cut Chinese firms off from Nvidia's most powerful hardware.
To give the pitch substance, China pledged 5,000 AI training opportunities for developing countries over an unspecified period and promised cooperation with ASEAN, the Arab League and the African Union. The conference itself, which Chinese officials said featured roughly 300 product debuts, doubled as a showcase for a domestic industry that Beijing insists is thriving despite the chip squeeze.
United Nations Secretary-General António Guterres attended and addressed the opening ceremony — lending the proceedings a stamp of multilateral legitimacy, even as he pointedly avoided endorsing any single bloc. "Technology that will shape the future of humanity must be shaped by all of humanity," Guterres said, warning that AI governance "cannot be governed by a handful of countries or companies" and that "every nation needs a seat at the table." He noted that "one-third of humanity is still offline," and cautioned that without deliberate action the technology could produce "greater inequalities, greater divides in income, in opportunity, in security."
Why it matters
Strip away the diplomatic gloss and what emerged in Shanghai is the clearest attempt yet to build a third pole in AI governance. The EU AI Act, now phasing into force, is a rules-first, risk-tiered regime built around binding obligations and heavy compliance. The US approach, reshaped by executive action under the current administration, leans toward light-touch domestic rules paired with aggressive control of the hardware supply chain — the chips, the tooling, the fabrication know-how. WAICO is pitched as neither: a development-first, sovereignty-friendly forum that promises access to technology and training without the conditionality of Brussels or the gatekeeping of Washington.
For many of the 29 founding members, that framing is genuinely attractive. Countries locked out of frontier compute by export controls, or wary of adopting EU-style rules they had no hand in writing, are being offered a seat, a training pipeline and a headquarters they can actually reach. Whether WAICO delivers infrastructure or merely a venue is the open question — an intergovernmental body with a Shanghai address and a founding declaration is a long way from a functioning standards-setter with enforcement teeth.
It is worth treating Beijing's numbers as claims rather than confirmed outcomes. The pledge of 5,000 training slots, the promised cooperation with regional blocs and the characterization of WAICO as a body "guided by the principles of the United Nations Charter" are all, for now, government statements of intent. The organization has no published charter detailing voting rights, funding or how — or whether — it will produce binding standards.
There is also a tension the summit did not resolve. Guterres used his platform to push the UN's own emerging architecture — the Global Digital Compact, the new Independent International Scientific Panel on AI, and a proposed Global Fund for AI — as the legitimate multilateral track. His presence in Shanghai can be read two ways: as tacit blessing of China's initiative, or as a reminder that the UN, not any national capital, intends to be the ultimate convener. More than 20 countries, including China, have already nominated centres for a UN-backed AI capacity-building network, a program that overlaps awkwardly with what WAICO now claims to offer.
What to watch next
The real test of WAICO will not be its launch communiqué but its first year of institutional plumbing: a published charter, a secretariat, a budget, and any concrete standard or training cohort that materializes. Watch whether additional countries — particularly fence-sitters in Southeast Asia, Africa and the Gulf — sign on, and whether any US ally defects toward the Shanghai body. Watch, too, for Washington's response; a governance bloc that explicitly frames export controls as illegitimate is unlikely to go unanswered. If WAICO becomes a durable institution rather than a summit set piece, July 17, 2026 will be remembered as the day AI governance formally fractured into competing camps.
"AI development should not be a solo performance by a single country, but a symphony of international cooperation."— Xi Jinping, President of China