Xi's WAICO message: AI as common goods, with Beijing setting the pace

Chinese President Xi Jinping took three slots at the World AI Conference in Shanghai on 17 July 2026 to make a single, repeatable point: artificial intelligence should be steered by "humanity's common values" and directed toward "shared prosperity and common security," not left to market gravity. The remarks, posted to X by CGTN at 15:15 UTC and again at 15:17 UTC, frame AI as a public good while leaving the operating definition to Chinese-led institutions. They are the most direct articulation yet of how Beijing intends to position itself as the convener, rather than the rule-taker, of the technology's diplomatic infrastructure.

What is being sold here is not a treaty. It is a vocabulary. The phrase "AI for good" has circulated for years; Beijing is now packaging it with a China-centred institutional layer (the World AI Conference, the AI Capacity-Building Action Plan, and a growing lattice of bilateral training programmes) that travels faster than any single piece of legislation can. The test is whether that vocabulary can outpace the European Union's risk-based rulebook and Washington's chip-and-licensing regime.

The shape of the Shanghai speech

Xi's three appearances on the X feed trace a deliberate arc. The first, at 15:15 UTC, set the tone: AI should be "a driver for shared prosperity and common security," language that mirrors earlier Chinese statements on cross-border data flows and the "community of common destiny." The second, two minutes later, deepened the philosophical claim, arguing that the values shaping AI must be "humanity's common values" rather than the preferences of any one jurisdiction. The third, posted at 15:53 UTC, was bilateral, recasting China as Thailand's "trusted strategic partner," a softer diplomatic register that signals the conference's audience is not only technologists in the plenary hall but cabinets in the Global South weighing whom to talk to first.

The arc matters. Public-goods framing lets Beijing argue against what it calls "technological monopolies" and "small-yard, high-fence" restrictions without naming the United States by name. Pairing that argument with a partnership offer to a Southeast Asian neighbour converts the rhetoric into a recruitment tool.

Why the framing lands where it does

The Chinese development model has measurable strengths that the speech does not invent: it builds infrastructure quickly, scales manufacturing capacity, and concentrates research funding in ways the Western system can no longer match without sustained political will. The country's leading AI laboratories, its domestic chip design houses, and its city-level compute clusters have all benefited from that pipeline. Pretending otherwise is not serious reporting.

What the speech does is attach that industrial record to a normative claim. By placing AI inside a "common values" register and by invoking "shared prosperity," Xi reaches past the United Nations General Assembly in New York to a wider audience of capitals that have concluded the existing multilateral system has under-delivered on their priorities. The audience is Brasilia, Jakarta, Pretoria, and Cairo, all of whom have heard similar language from Beijing over the past three years on climate finance and infrastructure.

The structural question is whether the rhetoric is matched by access. China's domestic compute buildout and its willingness to share models and training pipelines through programmes such as the AI Capacity-Building Action Plan have given Beijing something to offer beyond principles. The Western counter-offer, by contrast, is currently organised around export controls and ethics audits. Both are legitimate; they are not symmetric in what they hand a minister in Nairobi.

What it competes with, and where it does not

The European Union has spent the past three years constructing a risk-based regulatory architecture, which is real, binding, and has been the global reference point. The United States has tilted toward a chip-and-licensing regime built on industrial policy logic, more permissive than the EU's but narrower in where it engages. China's approach is the third pole: standards-led, narratively proactive, and willing to host summits.

None of the three has a working global compact. The EU's rules bind European companies. The U.S. rules bind U.S.-allied chip flows. The Chinese rules bind the deployment of AI inside Chinese jurisdiction, with voluntarist guidance for foreign partners. The gap between these three regulatory spheres is what Beijing is positioning itself to fill. WAICO is the venue.

The counter-reading is that "common values" is precisely the phrase that lets Beijing avoid committing to anything enforceable. An auditor in Brussels can read the EU AI Act and find a checklist. A regulator in Washington can read the Bureau of Industry and Security rules and find a part number. A regulator in Abuja, reading the Shanghai output, finds a declaration. That asymmetry will not be lost on attendees.

The stakes for the rest of 2026

The next four months will test whether the Shanghai speech becomes a working coalition or a photo opportunity. The signals to watch are concrete: whether the BRICS bloc (under its expanded 2024 membership) signs onto any common AI language, whether African Union digital ministers follow the Thai example of bilateral engagement with Beijing, and whether China's compute-export rules loosen enough to make the "shared prosperity" claim feel more than rhetorical.

On the other side, Washington's choices on licensing, and Brussels' choices on extraterritorial reach, will determine whether the Western pole can credibly compete for the same diplomatic ground. A rulebook that only its own companies have to follow is unlikely to win the Global South's loyalty; the EU has internalised that lesson on data flows and is now grappling with it on AI.

Two things remain genuinely uncertain. First, the sources do not yet disclose which bilateral agreements were signed on the WAICO margins; only the general framing of partnership is on the record. Second, no Western wire has published a direct rebuttal from Washington, Brussels, or London to Xi's specific phrasing, which makes the diplomatic response harder to read than the speech itself. Until those pieces are in the public domain, the assessment of this conference has to stop short of declaring a winner.

Desk note: this publication reads the X feed posted by CGTN on 17 July 2026 as a primary document and does not paraphrase Western wire framings of the speech, which had not been published at the time of writing. Where Western coverage arrives, we will update.

Wire provenance

This editorial synthesis draws on the following public wire/social posts:

• https://x.com/cgtnofficial/status/HNb2mTnbUAA9cWB
• https://x.com/cgtnofficial/status/HNb3UntawAAyL7F
• https://x.com/cgtnofficial/status/HNcGznzbsAA1EXs