Abstract: Artificial intelligence has been deeply integrated into various industries and realms of the economy and society, reshaping the mode of production and life and driving a leap in productivity as well as a deep change in production relations. As a new governance paradigm for addressing complexity and uncertainty, experimentalist governance takes experimentation as its core logic and possesses characteristics including multi-stakeholder collaboration, flexible and dynamic governance bases, cumulative and iterative governance processes, and controlled but innovative governance outcomes. To effectively balance technological innovation and risk mitigation, and promote the healthy and orderly development of AI in the direction of advancing the common well-being of humanity, China needs to establish an experimentalist governance system for AI. This involves setting up a tripartite system of governance featuring goals of sustainable innovation, controllable risks, and malleable values; improving a multi-stakeholder collaborative governance structure with government coordination, enterprise leadership, scientific research support, and public participation; and refining governance mechanisms of risk management, information communication, rights protection, and liability exemption.

Key Words: Artificial intelligence; Experimentalist governance; Complex adaptive systems; Human-centered approach; Risk management mechanisms