The World Artificial Intelligence Cooperation Organization (WAICO) is a new intergovernmental body headquartered in Shanghai, established on July 16, 2026, when 29 founding countries signed the agreement on the eve of the World AI Conference (WAIC). UN Secretary-General António Guterres attended. WAICO's stated goal is international cooperation and global governance on AI, including help for Global South countries to accelerate AI development.

Which countries signed WAICO?

The 29 founding members are: Algeria, Belarus, Brazil, Cambodia, Cameroon, China, Congo, Cuba, Ethiopia, Indonesia, Kazakhstan, Kenya, Kyrgyzstan, Laos, Lesotho, Malaysia, Mozambique, Myanmar, Nicaragua, Oman, Pakistan, Russia, Senegal, Serbia, South Africa, Tajikistan, Uzbekistan, Venezuela, and Zambia. Notably absent: the US, EU member states, UK, Japan, Australia, Canada, South Korea, and India.

How does WAICO compare to the UN's AI advisory bodies?

The UN AI Advisory Body (2023) and the July 2025 UN General Assembly resolution create a scientific panel and dialogue framework but no enforcement, no treaty status, and no headquarters. WAICO is a treaty-like intergovernmental organization with a Shanghai headquarters, a signed founding agreement, and formal member states. In practice, WAICO is a Chinese-led venue while the UN process remains a Western-anchored, consensus-driven track.

Does WAICO threaten the EU AI Act or US AI executive orders?

No direct threat — WAICO does not have jurisdiction over EU or US companies. But it splits the diplomatic AI-governance conversation into two tracks: a Western track (EU AI Act, US executive orders, G7 Hiroshima Process, UK AISI, OECD) and a China-led Global South track (WAICO). Multinationals now face potentially conflicting norms on model transparency, export controls, and safety standards depending on which market they serve.

Quick Answer

WAICO vs UN AI Body vs US/EU Governance (July 2026)

Published: July 17, 2026

WAICO vs UN AI vs US/EU Governance (July 2026)

On July 16, 2026, China formalised a rival AI-governance venue. Twenty-nine countries signed the founding agreement for the World Artificial Intelligence Cooperation Organization (WAICO) in Shanghai, on the eve of the WAIC 2026 conference. UN Secretary-General António Guterres attended. President Xi Jinping opened WAIC on July 17 with a keynote calling for stepped-up global AI cooperation.

The AI-governance map now has two clear poles: a Western track (EU AI Act, US executive orders, G7 Hiroshima Process, UK AISI) and a China-led Global South track (WAICO). Here is how they compare.

Last verified: July 17, 2026

The Four Frameworks at a Glance

Framework	Origin	Legal Status	Enforcement	Coverage
WAICO	China + 28 signatories, July 16, 2026	Intergovernmental treaty organization	None yet; secretariat forming	Shanghai HQ; 29 member states
UN AI process	UN General Assembly, 2023-2025	Non-binding resolutions	None; advisory only	Global; all UN member states
EU AI Act	European Union, 2024	Binding EU regulation	Fines up to 7% of global turnover	EU market; extraterritorial reach
US framework	Executive orders + NIST + AISI	Executive + voluntary	Sectoral (FTC, DOJ, sector regulators)	US market + federal procurement

WAICO: What It Is

Headquartered in Shanghai — first major intergovernmental AI body outside the West.
Founding treaty signed July 16, 2026 — 29 states, mostly Global South plus Russia and Belarus.
Stated purpose: international cooperation on AI, “bridge the AI divide,” help Global South countries build AI capacity, and articulate a governance approach distinct from EU and US frameworks.
Announced programmes: 5,000 AI training and seminar opportunities for developing-country participants, cooperation with ASEAN and other blocs, a “China Wisdom for the World” AI case collection covering over 20 countries.
Political framing: positioned by Xi as an alternative to what China calls “AI hegemony” and “security overreach” — a clear reference to US chip export controls and EU AI Act extraterritoriality.

What WAICO Is NOT (Yet)

Not a regulator. No fines, no model registrations, no evaluation regime.
Not a standards body. ISO/IEC JTC 1/SC 42 still owns most technical AI standards work.
Not a treaty on model safety in the sense of the Bletchley/Seoul/Paris AI Safety Summit track.
Not universal. Absent members include: US, all 27 EU states, UK, Japan, Australia, Canada, South Korea, India, Singapore, Israel.

UN AI Process: The Existing Multilateral Track

Since 2023, the UN has built out its AI machinery:

UN AI Advisory Body (2023) — 39 experts, published “Governing AI for Humanity” report.
UN Global Digital Compact (September 2024) — commitments on AI safety, capacity-building, data governance.
UN General Assembly resolution on AI (March 2024, updated July 2025) — non-binding but universally signed.
UN Scientific Panel on AI (established 2025) — international scientific body modelled loosely on IPCC.

Limits: no treaty, no enforcement, no headquarters, no budget commensurate with the task. WAICO steps into that vacuum with an actual building, staff, and a founding treaty — from a Chinese-led coalition.

EU AI Act: The Regulatory Heavyweight

In force since 2024, phased application through 2026-2027.
Risk-based framework: unacceptable risk (banned), high risk (heavy obligations), limited risk (transparency), minimal risk (free).
GPAI obligations: general-purpose AI providers face documentation, evaluation, and transparency requirements; providers of “systemic-risk” GPAI models (>10^25 FLOPs training) face additional cybersecurity and safety obligations.
Fines: up to €35M or 7% of global turnover — the largest AI-specific penalties anywhere.
Extraterritorial reach: applies to any provider placing AI on the EU market. This is why every major US and Chinese lab has EU compliance teams.

US Framework: Sectoral + Executive

Trump-era AI executive orders (2025-2026) rolled back the Biden 2023 EO and replaced it with a lighter-touch, competition-first framework.
NIST AI Risk Management Framework — voluntary, widely used.
US AI Safety Institute (AISI) — federal evaluation body; still funded and operating as of July 2026 despite political turbulence.
Sector regulators — FTC on consumer protection, SEC on AI disclosure, DOJ on antitrust, sector-specific rules for finance, healthcare, autonomous vehicles.
Export controls — the real teeth: chip export controls to China (2022 → 2023 → 2024 → 2025 tightening cycles) remain the strongest US governance lever.

The Fault Lines WAICO Highlights

WAICO’s July 16 signing crystallises three fault lines that will shape AI policy for the rest of the decade:

1. Chip access. WAICO members are largely on the wrong side of US export controls. Xi explicitly criticised “security overreach” — code for US chip restrictions.

2. Model transparency. EU AI Act demands training-data documentation, safety evaluations, and post-market monitoring. Chinese-led governance tends to emphasise state approval processes over transparency to the market.

3. Global South representation. The UN track is slow. The EU and US frameworks apply to their own markets. WAICO offers Global South states a venue where they are founding members, not observers.

What This Means for Companies

Multinationals: you now face potentially divergent norms depending on market. An EU AI Act-compliant frontier model may still face additional Chinese state approval. A WAICO-endorsed governance framework may not satisfy EU high-risk obligations.

Frontier labs: the diplomatic track is fragmenting. Expect distinct policy engagement teams for the EU, US, UK, and now the WAICO bloc.

Global South AI startups: WAICO offers scholarships, capacity-building programmes, and potentially cloud/compute access via Chinese providers. Real economic value for these markets.

Enterprises in emerging markets: if your government joined WAICO, expect Chinese cloud, Chinese models, and Chinese standards to be pushed as the “sovereign” alternative to US hyperscalers.

What to Watch Next

First WAICO working groups — expected within 6 months. Watch which topics are prioritised: safety, capacity, standards, or export policy.
UN AI Scientific Panel first report — could compete with WAICO’s technical positioning.
EU AI Act GPAI compliance — major frontier labs face August 2026 GPAI code-of-practice deadlines.
US-China dialogue — whether WAICO becomes a venue for direct US-China AI talks (unlikely given US absence) or a purely China-led forum.
India’s position — India is absent from WAICO. Its GPAI initiative and IndiaAI Mission put it in a swing role between the two tracks.

Bottom Line

WAICO is not a substitute for the EU AI Act, the US framework, or the UN process — it is an alternative diplomatic venue for the Global South plus Russia. For companies, the practical impact in the next 12 months is limited: no fines, no registrations, no compliance obligations from WAICO itself. But strategically, July 16-17, 2026 marks the moment where AI governance stopped being a Western-led conversation and became a two-track world.

Watch what WAICO actually does over the next year — a treaty is only as strong as its follow-through.

Sources

WAICO founding agreement coverage: Global Times, Anadolu
WAIC 2026 official portal: shanghai.gov.cn WAIC 2026
Xi keynote coverage: CNBC
EU AI Act: European Commission AI Act page