摘要
在人工智能时代,有必要甄别何者才是刑法面临的真正挑战,而不能动辄以新技术、新风险为由建造新的智识系统。人工智能只是人类为自身之目的而研发的高级工具,它无法被赋予“权利义务统一性”,难以改变二者之间的主客体关系。由于法人具有非生命体形象,学界常常借助单位犯罪类比论证人工智能的刑事责任主体地位,但二者存在根本不同的运行原理,肯定论对刑法中的“辨认控制能力”的认识也存在严重的以偏概全。惩罚人工智能不能实现报应、预防等目的,对其设定的刑罚引发了“技术失控-技术可控”“特殊预防无效-特殊预防有效”“刑罚设计-非刑罚性”“AI主体性-AI工具性”“消减风险-加剧风险”等五大悖论。在人工智能侵权案件中,犯罪主体只有自然人或单位,人类中心主义的责任体系具有恒久适应力,不能因应前沿科技而将刑法重构为技术管理法,更不能将之建立在修辞和想象之上。
In the era of Artificial Intelligence,it is necessary to identify what the real challenge is and not to build a new intellectual system based on new technologies and new risks. Artificial Intelli- gence is only a high-level tool developed by human beings for some purposes. It can not be given the “unity of rights and obligations”and it is difficult to change the relationship between the subject and ob- ject. Because corporations have non-living images,academic circles often use the analogy of corpora- tion crimes to demonstrate the criminal responsibility of AI. However,they have fundamentally different operating principles. The scholars keep serious misunderstanding of the“identification and control abili- ty”in criminal law. Punishing AI can not achieve the purpose of retribution,prevention. It has caused five major paradoxes. In the AI infringement cases,the criminal liability subject is only a natural person or unit. The responsibility system of anthropocentrism has long-term adaptability,which can’t be re- constructed into a technical management law in response to advancing technology,and it can not be built on rhetoric or imagination.
出处
《比较法研究》
CSSCI
北大核心
2019年第4期123-137,共15页
Journal of Comparative Law
基金
国家社会科学基金青年项目“网络时代刑法解释理念与方法研究”(18CFX042)的阶段性成果
中央高校基本科研业务费及江苏高校“青蓝工程”资助
关键词
人工智能
主客体关系
权利义务相统一
罪责能力
刑罚悖论
技术管理法
Artificial Intelligence
subject-object relationship
unity of rights and obligations
the liability
paradoxes of penalty
technique management law
