自动驾驶中的应急算法——一个对刑法学的挑战  

Contingency Algorithms in Automated Driving:a Challenge to Criminal Law

作　　者：弗兰克·彼得·舒斯特 张正昕(译) Frank Peter Schuster;ZHANG Zhengxin

机构地区：[1]德国维尔茨堡大学 [2]德国维尔茨堡大学法学院

出　　处：《中国政法大学学报》2021年第5期296-304,共9页Journal Of CUPL

摘　　要：在难以避免的交通事故中选择牺牲一部分人去拯救另一部分人,在刑法学的视角上存在着构成故意杀人罪的可能性。尽管汽车工业承诺此种情况几乎不会发生,但该问题依旧被证明是妨碍自动驾驶技术可接受性与创新性的巨大障碍。刑法传统理论一般将道德困境问题的出罪理由归结为违法层面的义务冲突理论与罪责层面的期待不可能理论,而自动驾驶的情形虽然与传统教科书案例存在一定差异,但在降低道路交通死亡总体人数这一道德目标的指引下,寻求该问题的法律解决方案仍然是可能实现的。Choosing to sacrifice some people to save others in an unavoidable traffic accident can constitute intentional homicide from a criminal law perspective.Despite the automobile industry’s promise that this will rarely happen,the problem continues to prove to be a huge obstacle to the acceptability and innovation of autonomous driving technology.Traditional theories of criminal law generally attribute the decriminalization of moral dilemmas to“conflicts of obligation”at the offense level and“unreasonable expectation”at the culpability level.Although the situation of autonomous driving differs somewhat from traditional textbook cases,it is still possible to find a legal solution to the problem under the guidance of the moral goal of reducing the overall number of road traffic fatalities.

关 键 词：扳道工案件 期待不可能 道德目标 义务冲突 

分 类 号：D920.5[政治法律—法学]