Zhou Xiaochuan, former governor of the People's Bank of China, stated on October 20 that artificial intelligence (AI) applications are capable of stimulating productivity growth. However, he warned about potential risks to financial stability associated with large-scale investments in AI infrastructure.
Speaking at the 'Band' Summit in Shanghai on October 20, 2026, Zhou noted that the rapid development of AI and the uncertainty surrounding its progress make it difficult to forecast the pace and scale of AI-driven productivity growth.
The summit, themed 'Technology, Geopolitics, and the World Economy,' is being held jointly by the China Finance Forum 40 and Tsinghua University in Shanghai from October 19 to 22.
Zhou identified three key areas for AI-related investments. The first is large models, as well as the computing power and data centers required to train them. The second area is the AI applications themselves. The third includes supporting industries such as semiconductors, electricity, and liquid cooling. He emphasized that the level of applications will be the main driver of productivity gains, noting the promise of lightweight models. Such models can embed AI capabilities into cars, laptops, and smartphones, expanding the practical application of the technology.
Furthermore, Zhou pointed to the significance of advancements in large models, memory chips, and large-scale simulation models as important technological breakthroughs. However, he questioned the degree of dependence of future AI models and applications on the computing power of data centers, given the enormous funds being directed toward building such centers.
Regarding international cooperation, Zhou stated that the far-reaching impact of AI on the global economy and society requires coordinated efforts, despite the fact that rapid development creates difficulties for governments in establishing regulatory frameworks. He highlighted three priorities for cooperation: controlling military applications of AI, combating fraud and hacking, and protecting personal privacy when using data. He added that broader implications for employment, income distribution, and social structures require further study.
Zhou also called for pricing mechanisms to play a more significant role in adapting economies to changes in global supply and demand, as well as in changing industrial structure, instead of over-relying on administrative measures. Meanwhile, he warned that failed investments in AI and related infrastructure could trigger financial turbulence or even a crisis, given the colossal sums being invested in this sector. He strongly urged decision-makers to conduct research and prepare contingency plans to support the economy and assist financial institutions in the event of severe market shocks.