Meng-Chun Liu, Director of the First Research Division at the Chung-Hua Institution for Economic Research (CIER), explains that the development of AI in China and the United States reflects a cyclical, interactive relationship. This dynamic is characterized by a process of “China catching up, the US containing, China leapfrogging, and global diffusion.” The pattern demonstrates that AI competition is not a straightforward confrontation but rather a dynamic equilibrium. This structure drives the global AI system toward fragmentation rather than convergence.
Director Meng-Chun Liu observes that China has adopted a leapfrog development strategy, integrating public and private sector resources to advance open innovation platforms while leveraging the RISC-V architecture and hardware alternatives to reduce dependence on high-end GPUs. Meanwhile, China is actively positioning itself in Global South markets with products that are “affordable and sufficiently good in quality,” demonstrating latecomer advantages and technological co-evolution characteristics.
Emerging economies are establishing autonomous AI systems, propelling the global AI ecosystem toward multipolarity. Future AI governance will evolve toward a model that is “fragmented yet interconnected.” Director Meng-Chun Liu emphasizes that to maintain a stable and sustainable technological order, countries must strengthen cooperation on interoperability, security, and data governance to build a balanced and inclusive new global AI order.
Author: CIER Editorial Team
Date: October 23, 2025