摘要
自动驾驶汽车的事故原因包括存在制造缺陷与存在设计缺陷两种,在因后者造成的事故中,制造商违反了注意义务,创设出了法不容许的风险,具有刑事违法性。自动驾驶汽车的使用者及制造商与法益损害均存在因果关系,基于信赖原则,使用者对“违反交通法规”不具有预见可能性,无须承担刑事责任。成立过失犯不要求行为人认识到具体的因果流程与致害行为,故只要制造商违反注意义务,创设出法不容许的风险,便对法益损害具有预见可能性,算法黑箱不能阻却其刑事责任。算法歧视有违法治国家平等保护之基本理念,在并未完全避免的情况下,原则上不应允许自动驾驶汽车上路。
The causes of autonomous vehicle accidents include manufacturing defects and design defects.In accidents resulting from manufacturing defects,the deservedness of constituent elements of a crime can be eliminated by“allowable risks”on condition that the relevant production standards are met.In contrast,in accidents resulting from design defects,the manufacturer violates the duty of care and creates unallowable risks and therefore should be held criminally responsible.The driving behavior of an autonomous car is jointly determined by the driving system and the user through the built-in program and the choice of driving route.The relationship between the two is that of“cumulative competition and cooperation”.Therefore,both the user and the manufacturer have a causal relationship with the damage to legal interests in a traffic accident.Based on the principle of“trust”,the user has no possibility of foreseeing a self-driving car's behavior of“violating traffic regulations”and causing a traffic accident and,according to the doctrine of accountability,needs not bear criminal responsibility for the damage results in principle.However,if a user knows that a self-driving car violates transportation management regulations but still rides it on the road,thus causing a traffic accident,he should be held criminally responsible for the traffic accident.According to the relevant provisions on artificial intelligence ethics,the manufacturer has the obligation to reasonably explain the operation principle of its vehicles.If it cannot clearly explain the operation principle of the autonomous vehicle due to the“algorithmic black box”,it violates the duty of care by making it impossible to determine the behavior of the system.The establishment of a negligent offense does not require the actor to recognize the specific causal process and harmful behavior,as long as it is known that his behavior has created an unacceptable risk.Therefore,as long as a manufacturer violates the duty of care and creates a risk not allowed by law,it should be regarded as being able to predict the legal interest damage caused by the vehicle and bear the corresponding criminal responsibility,and it can not be exonerated on the ground that it is unable to foresee the specific harmful behavior because of the existence of the“algorithmic black box”.For criminal policy,criminal law is an insurmountable barrier and the content of criminal law provisions cannot be arbitrarily created or removed by means of criminal policy or civil agreement.The belief that imposing criminal liability on the manufacturer for traffic accidents caused by autonomous vehicles will“suppress technology”and“hinder the development of science and technology”is a pure value judgment that lacks a reasonable theoretical basis and therefore cannot be the basis for decriminalization.“Algorithmic discrimination”will turn some people into“abandoned children”of science and technology development,which will not only cause the public to completely lose confidence in law but also violate the basic concept of equal protection by law.In principle,self-driving cars should not be allowed on the road before the relevant issues are resolved.
出处
《环球法律评论》
CSSCI
北大核心
2024年第5期56-71,共16页
Global Law Review
基金
2024年度四川省犯罪防控研究中心重点项目“行政犯时代违法性认识错误出罪路径研究”(FZFK24-03)的研究成果。