中图分类号：D92
文献标识码：A
DOI:10.19358/j.issn.2097-1788.2023.03.007
引用格式：韦欣荃.自动驾驶极端情况下的算法治理探究［J］.网络安全与数据治理，2023,42(3):38-45.

Research on algorithm governance in extreme situations of autonomous driving

Abstract： With the development of automated and connected vehicles, accidentalgorithms in extreme traffic situations have transformed traditional trolley problem into realworld dilemmas. Because accidentalgorithms are directly related to the safety of all road traffic users, and also involve the allocation of responsibility for traffic accidents, it is important to find a feasible solution before the industry of automated and connected vehicles is officially commercialized. There are several kinds of approaches to deal with this problem, like applying ethics theory, coding ethics with machine learning and presetting users’ moral choice. Based on the advantages and disadvantages of these existing approaches, this paper proposes a possible risk allocation plan, setting certain rules to exclude producers’ criminal liability and alleviate civil liability, hoping to build a regulatory framework in extreme conditions with the combination of law, ethics and technology.

Key words : automated and connected vehicles；accidentalgorithms；risk allocation