On Legal Risk Regulation of Artificial Intelligence Systems-A Case Study of the EU Artificial Intelligence Act

The European Union’s “Artificial Intelligence Act” (hereinafter referred to as the “Artificial Intelligence Act” or “this Act”) serves as the world’s first technological law that conducts comprehensive supervision of the risks of artificial intelligence. It offers novel ideas for global artificial intelligence legislation, emerging as a pioneering norm in the global legal supervision of artificial intelligence and presenting reference significance. The underlying regulatory logic of the “Artificial Intelligence Act” lies in the systematic regulation of technological risks, that is, with risk supervision as the core approach, after making a comprehensive judgment by considering factors such as the concept of risk, risk levels, and assessment difficulty, extensive intervention is carried out from aspects such as the development direction of artificial intelligence systems, prohibitive principles, the framework of rights and obligations, and damage remedy measures. For other countries, it is imperative to summarize and appropriately draw upon the legislative experience of the “Artificial Intelligence Act” and advance the progress of risk regulation of artificial intelligence systems. Chinese legislative workers can, on the basis of sorting out and summarizing the systematic legislative experience in the field of artificial intelligence governance of the European Union, conduct targeted innovation and comprehensive adaptation to China’s artificial intelligence legislation, and establish a comprehensive risk evaluation standard for artificial intelligence, consolidate rights and obligations in the artificial intelligence value chain, and promote the governance of artificial intelligence in China towards the direction of forward-looking risk, detailed responsibility, and clear regulation.