• At the World AI Conference in Shanghai, Xi Jinping framed AI as a matter for global cooperation, not a "solo performance by any single country," while criticizing the "overstretching" of national-security concepts—a clear swipe at U.S. export controls.
  • China launched the World AI Cooperation Organization (WAICO) with 29 countries, including Pakistan, Russia, and Kazakhstan, positioning it as a counterweight to U.S.-led tech alliances.
  • The initiative offers lower-cost, open-weight AI models and governance support to developing nations, intensifying competition over chips, standards, and diplomatic influence.

A New Bloc Takes Shape

China is making a calculated play to reshape the global AI order, casting itself as the champion of inclusive, affordable technology while presenting the United States as the gatekeeper of an exclusive club. At the July 2026 World Artificial Intelligence Conference (WAIC) in Shanghai, President Xi Jinping declared that AI development should not be "a solo performance by any single country," urging international collaboration and warning against the "overstretching" of national-security justifications for technology restrictions—a thinly veiled reference to Washington's export controls on advanced semiconductors to China.

The centerpiece of the announcement: the World AI Cooperation Organization (WAICO), a Shanghai-headquartered body with 29 founding members, including Pakistan, Russia, and Kazakhstan. Beijing is pitching WAICO as a forum for inclusive AI governance and capacity-building. Analysts see it as a direct answer to Pax Silica, the U.S.-led framework for coordinating AI partnerships among allied nations, which reportedly has 35 participating countries.

Competing Visions of AI Governance

The two governance models are now on a collision course. China's approach emphasizes state-to-state cooperation, technological sovereignty, and open-source or open-weight model availability. Xi pledged expanded AI cooperation with ASEAN, the African Union, Arab states, Latin American and Caribbean nations, the Shanghai Cooperation Organization, and BRICS. Specific commitments include 5,000 AI training slots for developing-country participants over five years and access for 30 countries to a Chinese AI weather early-warning tool.

The U.S. model, by contrast, prioritizes innovation leadership, secure supply chains, and allied coordination—backed by chip export controls that Washington treats as essential to national security. Beijing frames those same controls as an effort to preserve an American AI monopoly.

"What institutional investors like us are really focused on is regulatory stability," said a person familiar with the discussions, though the remark was made in a different context about private markets. The sentiment nonetheless underscores the broader industry appetite for predictability—something neither bloc has yet delivered.

A Market for Cheaper Models

China's push is not just diplomatic; it is commercial. Chinese labs are aggressively promoting lower-cost, open-weight models such as DeepSeek, Moonshot AI's Kimi, and Z.ai's GLM series. These systems are cheaper to deploy than many U.S. proprietary offerings and have particular appeal in developing economies where frontier U.S. systems are expensive or subject to use restrictions.

Industry trendlines show Chinese models narrowing the gap with leading U.S. systems on some benchmarks, though restrictions on advanced chips may constrain the scale of computing capacity available to Chinese providers. Leading U.S. labs like OpenAI and Anthropic continue to rely on closed-model approaches, retaining tight control over their most capable systems.

For businesses and developers, the emerging landscape offers more choices—but also higher compliance burdens, divergent regulations, and the risk of incompatible technology stacks. The bifurcation of chip supply chains, cloud services, and technical standards could add costs and complicate cross-border commerce even as it widens access to AI tools.

Diplomacy on the Edge

The issue has moved into high-level diplomacy. Ahead of an expected Trump–Xi meeting this week, U.S. Treasury Secretary Scott Bessent said Washington and Beijing have discussed an AI dialogue mechanism and a possible notification system for AI incidents that rise to a national-security level. Such a channel would be modest but meaningful, creating a way to communicate during a high-stakes AI-related crisis.

Both countries face shared threats from AI-enabled cyberattacks, misuse in biology, and failures of powerful AI systems. Xi himself stressed human control, early-warning systems, and emergency response—signaling that Beijing is publicly elevating these concerns, even as it criticizes U.S. security justifications.

What to Watch

In the near term, expect continued friction over advanced chips, cloud computing, model access, and alleged model distillation—claims that Chinese labs extract capabilities from American models, which Beijing disputes. China is likely to deepen WAICO, BRICS, and Global South programs with training, deployments, and standards initiatives. A limited bilateral safety arrangement remains plausible because it serves both sides without requiring either to concede strategic advantage.

Over the longer term, the most likely outcome is managed rivalry, not a decisive winner. The U.S. may retain advantages in high-end chips, frontier labs, and capital markets, while China strengthens its position in efficient open-weight models, domestic deployment, and AI partnerships across developing economies. The key uncertainty is whether competition fragments into a "two-stack" AI world—or whether shared risks force practical coordination.

Spokespeople for the U.S. Treasury and China's Ministry of Foreign Affairs did not immediately respond to requests for comment.

Correction: An earlier version of this article misstated the number of countries participating in the U.S.-led Pax Silica framework. It is 35, not 50.