Political risk / Jul 17, 2026 / 4 min
The Global South Just Got an AI Parliament
On July 16, 29 countries signed the World Artificial Intelligence Cooperation Organization into existence in Shanghai — giving Beijing a permanent intergovernmental seat to write AI rules for the developing world while Washington's 35-nation Pax Silica bloc locks up chips and the EU AI Act governs the West.
On July 16, representatives from 29 countries signed an agreement in Shanghai establishing the World Artificial Intelligence Cooperation Organization — an independent intergovernmental body with permanent headquarters in the city — and on July 17 Xi Jinping called it an "important milestone in the history of AI development," giving Beijing institutional machinery to write AI rules for the Global South while Washington's Pax Silica bloc and the EU AI Act govern everyone else.
What landed:
- Xinhua reported the July 16 signing made WAICO an "independent intergovernmental international organization headquartered in Shanghai."
- Chinese Foreign Minister Wang Yi signed for Beijing. Founding members include Brazil, Indonesia, Kazakhstan, Pakistan, Russia, and South Africa — but not the U.S., EU, UK, Japan, or South Korea, per CIO.
- UN Secretary-General António Guterres attended the ceremony. His presence lent weight — it is not a UN endorsement of China's agenda, as Telegraph.com noted.
- China first proposed WAICO in July 2025. Thursday's signatures turned a speech into a treaty.
Why this isn't just another summit:
- WAICO is the world's first intergovernmental organization devoted specifically to AI, per Telegraph.com — not a conference, not a voluntary framework.
- Researchers cited by CIO say it differs from Western initiatives in three ways: open membership to any sovereign state, no regime-type test, and an agenda built around the global capability divide.
- Countries that align with WAICO can write AI regulations incompatible with the EU AI Act, OECD principles, or the G7 Hiroshima Process, per TechTimes — a compliance fork for any enterprise operating across both blocs.
Xi's doctrine — July 17:
- "The establishment of the WAICO... will be an important milestone in the history of AI development," Xi said at WAIC's opening ceremony, per China Daily.
- He pledged 5,000 AI training and seminar slots for developing countries over the next five years and promised international application centers with ASEAN, the African Union, BRICS, and the Shanghai Cooperation Organization.
- Reuters reported Xi urged nations to seize the "rare and historic opportunity" of open-source AI and warned against "new historical injustices" from unequal access.
- Xi also warned against "overstretching the national security concept in the field of AI" — language aimed at U.S. chip export controls, per TechTimes.
The Western counterweight:
- Washington has 35 countries behind its Joint Statement on AI Opportunity, signed at the second Pax Silica Summit in June, per the Economic Times — a pro-growth, trusted-supply-chain framework led by Under Secretary Jacob Helberg.
- India joined Pax Silica but did not sign WAICO, per Telegraph.com — a clean pick of the American stack over Shanghai's rules room.
- Kazakhstan is the only country publicly listed in both coalitions, per Reuters reporting cited by Quews — proof most governments will hedge, not choose.
- Brazil signed WAICO while remaining in BRICS — another hedge, not a sealed bloc.
Who actually signed — all 29:
- 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, per Wikipedia's founding-member list compiled from the agreement.
What WAICO does not have yet:
- Telegraph.com cautions there is no disclosed WAICO budget, no enforcement mechanism, and no binding technology-transfer commitments — yet.
- The July 16 agreement is a founding charter, not proof the world has split into two sealed AI blocs.
- The tests ahead: ratification, secretariat control, whether standards enter national procurement law, and whether application centers run Chinese systems exclusively.
Why institutions may outlast models:
- Telegraph.com's core argument: most governments will never train a frontier model. Their question is who helps them deploy AI — and open-weight models like Kimi K3 make Chinese stacks cheaper to adopt locally.
- Engineers trained on Chinese systems, procurement built around Chinese standards, and ministries regulating Chinese terminology create ecosystem lock-in that survives any single benchmark crown.
- Competing governance bodies risk fractured safety research, incompatible agent standards, and an adoption arms race where safety becomes a commercial disadvantage.
What to watch:
- Whether additional countries sign the open agreement in Beijing.
- U.S.-China government-level AI talks under Trump — WAIC and WAICO landed the week before them.
- Whether WAICO standards show up in member-state procurement within 12 months.
- Enterprise compliance teams mapping dual-stack obligations across WAICO-aligned and EU/Western jurisdictions.
Convina's view: Kimi K3 was this week's fireworks — WAICO is the permanent address behind them. Beijing spent year one building models that work despite export controls; year two is building governments that standardize around them. Washington is not wrong that WAICO doubles as Chinese industrial diplomacy — training slots and application centers are customer acquisition for a Shanghai-based stack. But the U.S. reply is still mostly chips, clouds, and guest lists, while a billion people live in countries that just signed a charter saying AI governance should not require American permission. The world has not split yet. It now has two rulebooks and a race to see which one becomes ordinary outside the G7.