江汉学术 ›› 2025, Vol. 44 ›› Issue (1): 107-117.doi: 10.16388/j.cnki.cn42-1843/c.2025.01.010

• 法律专栏_法治中国建设的新视域与新话语 • 上一篇    

人工智能侵权责任主体认定——以自动驾驶为例

张家骥1,铁伊然1,周 鑫2   

  1. 1北京航空航天大学 法学院,北京 100191;2廊坊师范学院 社会发展学院,河北 廊坊 065000
  • 收稿日期:2024-07-18 出版日期:2025-02-15 发布日期:2025-01-08
  • 作者简介:张家骥,男,北京人,北京航空航天大学法学院副教授,博士,E-mail:jiajizh@hotmail.com;铁伊然,女,北京人,北京航空航天大学法学院学生,E-mail:bj061625@163.com;周 鑫,男,江苏连云港人,廊坊师范学院社会发展学院副教授,博士,E-mail:895331361@qq.com。
  • 基金资助:
    北京市大学生创新训练项目“从法哲学到实在法:人工智能侵权责任归责方案的完善——以无人驾驶领域为例”(S20231006316)

Identification of the Subject of AI Infringement Liability:Taking Autonomous Driving as Example

ZHANG Jiaji1,TIE Yiran1,ZHOU Xin2   

  1. 1Law School,Beihang University,Beijing 100191;2School of Social Development,Langfang Normal University,Langfang Hebei 065000
  • Received:2024-07-18 Online:2025-02-15 Published:2025-01-08

摘要: 人工智能可否享有法律主体地位是解决人工智能侵权责任归责主体认定这一现实问题的根源问题。从法哲学角度切入,创造性建构法律主体的阶层式标准体系及其应然条件,在此基础上提出赋予法律主体地位的阶层式判断体系。从抽象到具体,运用该体系作为判断标准,就自动驾驶人工智能侵权责任归责主体问题展开研究。弱人工智能不能成为法律主体,提出建立在生产方强制保险机制上的法律地位代表说。强人工智能根据自我意识的根源不同分为“机器伦理”式和“人类情感”式两类——前者不能成为法律主体,但人类设计的“机器伦理”将有效减少致损事件发生的概率,法律地位代表说依然适用;后者不应成为法律主体,否则将不满足价值判断,甚至是无意义的讨论。从理论到实践,在人工智能不作为法律主体的前提下,基于我国现行法与机动车责任保险机制,结合域外立法经验与案例,明确自动驾驶人工智能侵权责任归责主体,进而完善具有中国特色的自动驾驶领域人工智能侵权责任归责方案。

关键词: 人工智能侵权, 自动驾驶, 责任主体, 阶层式体系, 机器伦理

Abstract: Whether artificial intelligence(AI)can enjoy the status of legal subject is the root of solving the realistic problem of ascribing the subject of liability for AI tort liability. From the angle of legal philosophy, the paper creatively constructs the hierarchical standard system of a legal subject and its contingent conditions,and then puts forward the hierarchical judgment system of entrusting the legal subject status. From the abstract to the concrete,the problem of ascribing the subject of liability for AI tort liability is studied by using this system as the judgment standard. Since weak artificial intelligence cannot become a legal subject,the idea of legal status representative is put forward on the basis of the compulsory insurance mechanism of the production side. Strong artificial intelligence,according to the different sources of its self-consciousness,is divided into“machine ethics”and“human emotion.”The former cannot become a legal subject,but the human-designed“machine ethics”can effectively reduce the probability of damaging events,and therefore,the idea of legal status representative still works;the latter should not become a legal subject,for otherwise,it will not satisfy the value judgment,or even relevant discussion will be meaningless. Under the premise that artificial intelligence is not regarded as a legal subject,and based on China’s current law and motor vehicle liability insurance mechanism,the ascription of the subject of liability for AI tort liability is clearly defined from theory to practice,drawing on foreign legislative experience and cases. It helps to improve the scheme of ascribing the subject of liability for AI tort liability in the field of automatic driving with Chinese characteristics.

Key words: AI infringement, automatic driving, subject of liability, hierarchical system, machine ethics

中图分类号: