自动驾驶事故中的过失犯罪分析
作者:
中图分类号:

D914

基金项目:

国家社会科学基金青年项目“数字化时代网络犯罪罪量认定研究”(24CFX093);上海市哲学社会科学规划课题“资本市场改革背景下证券共同犯罪认定研究”(2022EFX005)


The analysis of negligent crimes in automatic driving accidents
Author:
  • 摘要
  • | |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • | |
  • 文章评论
    摘要:

    人工智能技术的快速发展和自动驾驶事故的发生引发了关于自动驾驶事故中刑事责任问题的讨论,其中涉及刑事责任主体、过失犯罪本质以及注意义务等内容,值得深入研究。自动驾驶技术可以分为简单自动化、部分自动化、高度自动化和完全自动化。在简单自动化以及部分自动化情形中,自动驾驶汽车交通事故的刑事责任承担与普通交通事故并无二致。在高度自动化和完全自动化的情况下,自动驾驶汽车的驾驶员因被限制参与驾驶操作而成为产品的使用者。自动驾驶汽车作为弱人工智能产品尚不能成为刑事责任主体。自动驾驶汽车的使用者不能完全免除道路交通安全责任。自动驾驶汽车的研发者需要承担产品刑事责任,此时的产品刑事责任包含道路交通安全责任。在我国刑法体系中,过失和故意是平行的两种主观心理态度,新过失论对过失犯罪的理解不符合现有刑法规定。过失犯罪并非旧过失论所言只考虑抽象的预见可能性。我国刑法规定中过于自信的过失是行为人在已经预见结果的情况下没有履行结果回避义务,疏忽大意的过失是行为人在具有预见可能性的情况下没有履行结果预见义务。自动驾驶事故中,如果自动驾驶汽车的使用者违反配合与保障自动驾驶系统正常工作的结果回避义务致使危害结果发生成立过失犯罪。如果使用者不负有操作义务则无需承担刑事责任。使用者的注意义务源于前置法,但不是所有前置法规定的义务都能成为过失犯罪注意义务。如果自动驾驶汽车的研发者违反保障道路交通安全的结果预见义务,致使危害结果发生,需要承担交通过失责任。自动驾驶系统在投入使用后并未脱离研发者的控制,因而自动驾驶汽车研发者承担的产品责任性质不同于普通汽车研发者,原有汽车驾驶者承担的交通过失责任转移至研发者。虽然结果预见义务是主观的,但是在实际判断自动驾驶汽车研发者是否违反结果预见义务的过程中需要相对客观的标准作为参考,即研发自动驾驶汽车当时的科学技术水平。自动驾驶系统存在算法黑箱不应成为否定研发者结果预见义务或结果回避义务的理由。

    Abstract:

    The rapid development of artificial intelligence technology and the occurrence of automatic driving accidents have led to a discussion on the issue of criminal liability in automatic driving accidents, involving the subject of criminal liability, the nature of the crime of negligence, and the duty of care, which is worthy of in-depth study. Automatic driving technology can be classified as simple automation, partial automation, high automation, and full automation. In the case of simple automation and partial automation, the criminal liability for traffic accidents involving self-driving cars is not different from that of ordinary traffic accidents, and there is no need for a separate discussion. In the case of high automation and full automation, the driver of a self-driving car becomes a user of the product because he is restricted from participating in driving operations. As a weak artificial intelligence product, self-driving cars cannot become the subject of criminal liability. Users of self-driving cars are not completely exempt from road traffic safety responsibilities. The developer of self-driving cars needs to bear the criminal liability of the product, and the criminal liability of the product in this case includes the responsibility for road traffic safety. Under China’s criminal law system, negligence and intent are two parallel subjective psychological attitudes, and the new negligence theory’s understanding of negligence crime does not conform to the existing criminal law provisions. The crime of negligence is not an abstract foreseeable possibility as the old negligence theory suggests. The overconfident negligence in China’s criminal law is the actor’s failure to perform the obligation to avoid the result when he foresees the result, and the negligence is the actor’s failure to perform the obligation to foresee the result when there is a possibility of foreseeing. In an automatic driving accident, if the user violates the obligation to cooperate with and ensure the normal operation of the self-driving system, resulting in the occurrence of harmful results, the crime of negligence is established. If the user is not obligated to operate, he or she is not subject to criminal liability. The user’s duty of care is derived from the precedent law, but not all the obligations stipulated in the precedent law can become the duty of care for negligence crimes. If the developer violates the obligation to foresee the consequences of ensuring road traffic safety, resulting in the occurrence of harmful results, it needs to bear the responsibility for the accident. The self-driving system has not been out of the control of the developer after it has been put into use, so the nature of product liability borne by the developer of the self-driving car is different from that of the ordinary car developer, and the liability for the accident of the original car driver is transferred to the developer. Although the obligation to foresee the results is subjective, in the process of actually judging whether the developer has violated the obligation to foresee the results, a relatively objective standard is required as a reference, that is, the scientific and technological capabilities at the time of the development of the self-driving car. The existence of algorithmic black boxes in automatic driving systems should not be a reason to deny the developer’s obligation to foresee or avoid results.

    参考文献
    相似文献
    引证文献
引用本文

林雨佳.自动驾驶事故中的过失犯罪分析[J].重庆大学学报社会科学版,2024,30(6):238-249. DOI:10.11835/j. issn.1008-5831. fx.2024.05.01

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 在线发布日期: 2025-02-13
文章二维码