From AI Plan to AI Practice: A Study of AI innovation in China’s Innovation System

Abstract

China has emerged as a global powerhouse in artificial intelligence (AI) innovation, and the Next Generation AI Development Plan (NGAIDP) signals the Chinese Communist Party’s (CCP) intent to lead in AI globally. At the same time, AI innovation in China unfolds in an institutional environment distinctly different from the West. As innovation system (IS) literature is grounded in market-driven contexts, there exits a need for understanding how non-Western, state-driven ISs operate, how centralized state control and market forces interact, and how AI innovation is fostered in this context. To address this gap and understand how China’s AI ambitions are orchestrated, the study asks “How is the CCP’s AI innovation strategy formulated-, influenced by-, and operationalized in China's Innovation Systems?”. It utilizes an embedded case study approach grounded in critical realism, to analyze cause-effect patterns and the interplay between China’s IS, three central AI plans, and five embedded cases: (1) China’s AI Talent, Education and Research System, (2) Baidu and Autonomous Vehicle AI, (3) Alibaba and Smart City AI, (4) Tencent and Healthcare AI, and (5) The Shenzhen AI Pilot Zone. It employs a theoretical framework grounded in theories from pioneering IS scholars including Freeman, Lundvall and Nelson, to understand how concepts such as national idiosyncrasies, innovation policy and knowledge flows are fostered and reflected in China’s AI efforts and IS. Most notably, the study finds the AI innovation strategy to represent a dynamic blend of state and market forces, driven by an ideological foundation rooted in Socialism with Chinese Characteristics. The strategy and IS are mutually tied to geopolitics and national idiosyncrasies, shaped by the CCP’s commitment to Indigenous Innovation, centralized control and the ideological pursuit of a Community of Common Destiny. The strategy is techno-nationalistic, domestically innovation-centric, and manifested in national AI plans, which are distributed to the regional-level through AI pilot Zones, and the enterprise-level through AI open innovation platforms. The AI plans are identified as non-binding policy tools taking on a more coercive role, and shaped around a triple helix where the government exerts more control over academia and industry than the literature suggests. Yet, top-down innovation strategies are found to coexist with bottom-up innovation processes in China’s unique institutional context. Influenced by factors such as Regionally Decentralized Authoritarianism, guanxi and Guojin Mintui, the strategies’ formulation and operationalization was identified as increasingly market-oriented, with private enterprises enjoying policy- and governmental support traditionally reserved for state-owned-enterprises. The study contributes to a deeper understanding of the complex nature of China’s IS, and finds AI being used to fundamentally reshape industries and surrounding IS. Finally, it identifies a distinctive characteristic, wherein AI policy has been used to cultivate a new type of IS. Merging national, regional, state, and market forces, AI, and open-innovation to alter existing industries, the study terms this Industry IS with Chinese characteristics.