Janice R. Andrews
Progress and Perspective: Assessing China’s AI Development
Recent breakthroughs by DeepSeek, a Hangzhou tech firm, have marked a notable advancement in China’s artificial intelligence capabilities, yet the path to global AI leadership remains complex, according to prominent academic Zheng Yongnian. Speaking from his position as dean at the Chinese University of Hong Kong’s Shenzhen campus, Zheng emphasized the importance of measured optimism over nationalist fervor.
DeepSeek’s December release of the V3 language model, followed by the R1 inference model, demonstrated China’s ability to create cost-effective, open-source AI systems comparable to those from OpenAI and Google. While Chinese media celebrated these achievements as evidence of resilience against U.S. technology restrictions, Zheng highlighted persistent challenges.
Key obstacles include China’s continued dependence on foreign semiconductor technology and limitations in data infrastructure. Unlike the U.S., which benefits from global data access, China faces constraints due to internet restrictions and limited collaboration between government and commercial sectors.
Zheng proposed several strategies to enhance China’s competitive position, including reducing regulatory barriers that may drive talent abroad and implementing policies to promote open-source AI development. He argued that widespread access to AI models could accelerate technological progress.
The nation currently requires security assessments for public AI deployments, while comprehensive AI legislation remains pending amid industry debate. As global competitors advance their AI capabilities, China faces the challenge of balancing innovation with regulatory oversight to maintain its technological momentum.
