Cross the river by feeling the stones again? The co-evolution of China’s data governance and the emerging artificial intelligence industry

Authors: Zhen Yu*, Tsinghua University, Zheng Liang, Tsinghua University
Topics: Economic Geography, China
Keywords: Data governance, Artificial Intelligence, Institution, Emerging Industry
Session Type: Paper
Day: 4/6/2019
Start / End Time: 3:05 PM / 4:45 PM
Room: Blue Room, Omni, East
Presentation File: No File Uploaded

Unlike past industry development paths where China always strived to catch-up, China has been one of the leading players around the world in the rise of artificial intelligence (AI) industry. Many observers have attributed this to China’s huge domestic market and massive volume of data, but what has been neglected is China’s social and institutional contexts that give rise to the market expansion and data accumulation. AI industry differs from other emerging industries in that data is the key production factor, and institutions around data have a singularly important role to play in the growth of the AI industry. While AI remains a controversial technology in many countries and some of them have set strict rules on data ownership and use, Chinese society, in general, has a very optimistic attitude towards AI and shows a high level of tolerance (and even ignorance) to AI’s ethical risks. China’s policies at both the central and the local level have much focused on technological innovation and industrial application but have lax regulations on data protection, which facilitate the emergence of AI industry but also lead to many governance challenges in protecting citizens’ privacy and safety. Meanwhile, incremental institutional changes regarding data are being made to adapt to the development of the AI industry. It seems that China’s governance of AI industry still follows the philosophy of crossing the river by feeling the stones, and explicit rules only come after social problems arise, which in turn affect the industry’s further development.

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login