Xi’s AI Diplomacy Takes the Stage in Shanghai

Xi’s AI diplomacy has never had a bigger stage than the one waiting for him in Shanghai this week. China’s president is set to walk into the World AI Conference not as a name on an invite list, as premiers have done in years past, but as the man delivering the opening keynote himself, a first in the event’s eight-year history.

Beijing confirmed the move on Monday, with Xi opening the conference on July 17 after previously leaving hosting duties to his premier. The World Artificial Intelligence Conference, running through July 20 alongside a parallel High-Level Meeting on Global AI Governance, arrives at a moment when Washington has been accusing Beijing of copying American models even as China tightens access to its own best systems for outside users.

That tension is precisely the backdrop against which Xi is expected to speak. A foreign ministry spokesperson said Xi will use the opening ceremony to lay out China’s policies, positions and proposals on AI development and governance in systematic detail. It’s less a tech-conference cameo than a declaration of intent.

The World AI Cooperation Organisation Waits for Its Moment

The institutional centrepiece to watch is the World AI Cooperation Organisation, or WAICO, a body China first proposed at last year’s conference, though no country has yet formally signed on as a member. The organisation is expected to headquarter permanently in Shanghai, and analysts expect this year’s keynote to finally give the loosely-defined proposal some real shape, even though Beijing hasn’t confirmed exactly what that will look like.

The pitch to the rest of the world is fairly blunt. Where the US has built its AI governance around export bans and restricted-entity lists, China is offering membership, open-weight models and cheaper access instead, an offer aimed squarely at countries Washington hasn’t courted. Beijing is also expected to push its open-source models as an affordable alternative to Western systems, arguing they widen access to the technology rather than restrict it.

State media have been laying the rhetorical groundwork for days. A People’s Daily commentary published this week argued that AI development shouldn’t drift toward monopolies that wall themselves off, insisting the technology should ultimately serve humanity at large. It’s the same framing Xi struck back in January, when he likened AI to a transformation on the scale of the steam engine, part of Beijing’s broader bet that spreading AI through the economy is central to future growth and tech self-sufficiency.

Hardware on Display, Numbers on the Table

The policy talk will share the room with hardware. Huawei is set to give the public its first proper look at the Atlas 950 SuperPoD, a supercomputing cluster unveiled late last year and built for large-scale training and inference, capable of scaling up to 8,192 Ascend AI processor cards.

The economic stakes attached to all this are hard to overstate. PwC projects AI could add $15.7 trillion to global GDP by 2030, with Asia capturing close to half of that gain, while McKinsey estimates AI adoption alone could add up to $1.2 trillion a year to China’s economy through gains in manufacturing, finance, healthcare and logistics.

Shanghai itself isn’t an incidental backdrop. The city has spent a decade building itself into China’s AI capital through municipal funding, compute subsidies and a growing cluster of research labs, and landing a permanent international body there would put a physical address behind that ambition.

Whether Xi’s words translate into countries actually joining WAICO is the real test. For now, the conference has become something bigger than a trade show, a platform where China is trying to write the vocabulary of global AI governance before anyone else gets the chance.

This is a developing story and will be updated as the World AI Conference in Shanghai unfolds.