人工智能风险的契约观及契约化治理机制研究

The development and application of artificial intelligence technologies are a double - edged sword: while solving many economic and social problems, they also contain huge risks. Therefore, it’s urgent to discuss the risks and the governance of artificial intelligence. Based on the extant research, this paper first proposes that three main types of risks—technology risk, society risk, and management risk—may emerge in the development and application of artificial intelligence technologies. Second, based on contract theory, this paper analogically analyzes the transaction problems caused by contract characteristics and three types of artificial intelligence risks, and proposes a contractual view of artificial intelligence risks. Artificial intelligence risks are not only caused by technological underdevelopment, but also by the incompleteness, implicitness, and opacity of contracts. Third, based on the contract governance literature, this paper focuses on the design, implementation, and supervision of contracts in the context of artificial intelligence applications, and proposes contractual governance mechanisms for artificial intelligence risks. Finally, from the perspective of system design, this paper proposes a guarantee system. This research constructs a comprehensive contractual view of artificial intelligence risks, proposes contractual governance mechanisms, and establishes guarantee systems for artificial intelligence risks. It therefore provides a new theoretical perspective for a deeper understanding of artificial intelligence risks, and expands contract theory from explicating business relationships between economic agents to relationships between human and machine, and between machines, which enriches the application of contract theory.