文献类型：期刊文献

中文题名：人工智能侵权责任认定

英文题名：Identification of artificial intelligence infringement liability

作者：许辉猛[1];王飞翔[1]

第一作者：许辉猛

机构：[1]河南财经政法大学民商经济法学院

第一机构：河南财经政法大学民商经济法学院|河南财经政法大学法学院

年份：2018

卷号：20

期号：4

起止页码：56-63

中文期刊名：长安大学学报：社会科学版

收录：国家哲学社会科学学术期刊数据库

语种：中文

中文关键词：人工智能;归责研究;例证法;无人驾驶汽车;侵权责任;责任主体

外文关键词：artificial intelligence;accountability studies;exemplification;autonomous cars;infringement subject;infringement liability

摘要：各国法律在认定人工智能侵权主体方面缺乏适当的请求权基础,大陆法系和英美法系国家对物之客体局限性的表达也使得人工智能处于法律主体认定的模糊地带。为了厘清人工智能产品侵权的各主体之间关系以及确认人工智能是否具有主体资格,以无人驾驶汽车为例,从生理因素、社会因素和心理因素论证人工智能产品的物权属性,得出无人驾驶汽车的产品属性,进而确定责任主体:生产者、销售者、设计者以及操作者。以《中华人民共和国侵权责任法》(以下简称《侵权责任法》)和《中华人民共和国产品质量法》(以下简称《产品质量法》)为规范依据,对比生产者、销售者、设计者、操作者之间的关系,同时设定3个侵权场景即无人驾驶汽车和行人、无人驾驶汽车和普通汽车以及无人驾驶汽车与无人驾驶汽车来确定四者在人工智能产品侵权时的法律地位。在人工智能侵权场合将举证责任分配给侵权者,由其来证明自己无过错,一方面减轻了被侵权人的举证责任,另一方面也能够在侵权行为和结果之间成功搭建因果关系。结果表明:人工智能本身不能作为责任主体承担侵权责任,而是作为客体并适用规范客体的法律来分担侵权主体之间的责任,在侵权事实出现时,针对不同的主体运用不同的归责原则,使责任得以合理分配。
The laws of manycountries lack a proper basis for the right of claim in identifying the subject of artificial intelligence infringement. The limited expressionsfortbe objects in Continental law system and Anglo American law system also leaves artificial intelligence an ambiguously identifiedlegal subject. In order to clarify the relationship between different subjects in artificial intelligence product infringement cases and confirm whether artificial intelligence is qualified as a subject, we take autonomous carsas an example. The real right attribute of artificial intelligence products is defined by physical, social and psychological factors, from which we get the product nature of autonomous cars, thus defining the subjects of liability： producers, sellers, designers and operators. Based on the Infringement Liability Lazy of the People＇s Republic of China （hereafter referred to as ＂Infringement Liability Lazy＂） and Product Quality Lazy of the People＇s Republic of China （hereafter referred to as ＂Product Quality Law＂）, we compared the relationship among producers, sellers, designers and operators, andset three infringement scenarios： the autonomous car and pedestrians, the autonomous car and ordinary car, and the autonomous car and other autonomous cars toidentify the legal status of the four subjects in infringement of artificial intelligence products. In the case of artificial intelligence infringement, the burden of proof will be assigned to the infringementsubject, who will prove that there is no fault on its side. On the one hand, it elevates the burden of proof from the victim, and on the other hand, the causal relationship can be successfully established between the infringementand the results. The results show that artificial intelligence itself cannot bear the infringement liability as the liability subject, but can share the liability with the infringements subject as the object according to the law governing the objects. When the infringement facts emerge, different imputation principles are applied to different subjects, so that the liability can be reasonably distributed.