China Just Signed Up 29 Countries to an AI Body the US Isn’t In

On July 16, 2026, representatives from 29 countries gathered in Shanghai and signed the founding agreement for the World Artificial Intelligence Cooperation Organization, or WAICO — a new intergovernmental body headquartered in China and focused on AI governance. The signing happened the evening before the World Artificial Intelligence Conference opened in the same city, with UN Secretary-General António Guterres in attendance. The next day, Xi Jinping used his conference keynote to call for what he described as a “just and equitable” system of global AI governance.

The list of who signed is notable. So is the list of who didn’t.

Who’s in the room

WAICO’s 29 founding members lean heavily toward the Global South: Indonesia, Brazil, Malaysia, South Africa, Senegal, Pakistan, Kazakhstan, Laos, and Russia are among the names that show up across the reporting. China’s Foreign Minister Wang Yi represented the host country at the signing ceremony. The organization is structured as an independent intergovernmental body, open to any sovereign state, and framed around development-centric AI cooperation rather than restriction.

Who’s not in the room is the more interesting list: the United States, the European Union and its member states, the UK, Japan, and South Korea are all absent, according to reporting from Computerworld and CIO. This isn’t a case of an invitation being declined — it’s a body built, by design or by circumstance, around a bloc of countries that sit outside the existing Western-led AI governance apparatus: the EU’s AI Act, the OECD’s AI Principles, and the G7’s Hiroshima Process.

This isn’t China’s first pitch

WAICO didn’t appear out of nowhere. Xi first floated the idea of a global AI cooperation body in July 2025, then repeated the pitch at the APEC summit in October 2025. What changed between “proposal” and “29 signatures on a founding document” is a year of China positioning itself as the AI partner willing to work with countries that Western frameworks treat as compliance targets rather than co-authors. We’ve written before about the German media regulator move against Google’s AI Overviews as one example of how the EU’s approach to AI tends to run through liability and restriction — that’s a defensible regulatory philosophy, but it’s also exactly the framing China is positioning WAICO against.

The stated goals, per the founding materials, are cooperation on AI regulation, benefit-sharing, and safety across member states. That’s vague enough to mean almost anything in practice, which is normal for a body’s founding text — the OECD’s AI Principles and the G7’s Hiroshima Process both started similarly broad before hardening into specific requirements over time.

The skeptical read, and why it’s not the whole story

The obvious cynical take: this is a soft-power exercise, a way for China to build diplomatic goodwill and export its own regulatory preferences to 28 other countries at once, dressed up in cooperation language. French officials have made a version of this argument directly, describing WAICO as a risk of normalizing an alternative, more state-directed model of AI governance that could undermine the Hiroshima Process consensus. That’s a real concern, not a talking point — governance frameworks that get baked into 29 countries’ regulatory systems are hard to unwind later, regardless of what anyone initially intended them to do.

But the dismissive version of that read — “it’s just propaganda, ignore it” — doesn’t hold up either. Indonesia, Brazil, Malaysia, and South Africa aren’t puppet states signing whatever Beijing puts in front of them; they’re countries making a calculation that the existing AI governance conversation, run mostly by the US, EU, and G7, hasn’t been offering them much of a seat. A framework that explicitly targets Global South participation is filling a gap that the incumbent institutions left open, whether or not that was their intent.

Both things are true at once: WAICO is a vehicle for Chinese influence, and it’s also a response to a real absence in how AI governance has been organized so far. Treating it as pure geopolitical theater misses the second part; treating it as a neutral, altruistic cooperation body misses the first.

Why this matters if you’re not in any of these 29 countries

Two reasons. First, Western AI labs — Anthropic, OpenAI, Google DeepMind, and the rest — now face a market where a third of the world’s countries are potentially building AI rules under a framework those labs had no hand in shaping and aren’t members of. That’s a genuine strategic complication, not a scheduling inconvenience, for any company planning to operate in WAICO member states.

Second, and more durable: AI governance frameworks tend to outlast the news cycle that produced them. The EU’s AI Act didn’t disappear after its launch week, and there’s no reason to expect WAICO’s rules — once member states start actually writing them — to be any different. If two incompatible governance blocs solidify over the next few years, “which AI framework does this country follow” becomes a real variable in market access, the same way data-localization laws already are.

What to actually watch

The founding signatures are the easy part. The test is what WAICO’s members actually produce over the next year: does it converge on genuinely new rules, or mostly restate what member states were already doing domestically? Does the roster grow beyond 29, particularly toward larger Global South economies still on the fence? And does any G7 country’s position soften — the absence of Japan and South Korea specifically is worth watching, since both have deeper economic ties to China than the US or EU do. If you want a framework for weighing a story like this against the geopolitical spin coming from both directions, we’ve laid one out here.

Sources: Al Jazeera, Computerworld, CIO, Xinhua, PYMNTS, CNBC