人工智能自主性犯罪的法律应对:从惩罚转向预防  

Legal Response to AI Autonomous Crime:From Punishment to Prevention

作　　者：王超 WANG Chao

机构地区：[1]中国人民公安大学法学院

出　　处：《阅江学刊》2025年第2期86-96,173,共12页Yuejiang Academic Journal

摘　　要：人工智能犯罪绝非“未来性命题”,目前引发刑法学界较大争议的是人工智能自主性犯罪问题。对此,“事后惩罚”方案认为应当承认人工智能的刑法主体地位,并在此基础上重构刑罚体系,以应对人工智能自主性犯罪。“事后惩罚”方案立足于惩罚犯罪,缺乏现实基础,并不可取。因此,人工智能自主性犯罪应当从事后惩罚转向事前预防。“事前预防”方案的核心在于风险预防,其思路是通过民法、行政法、刑法等不同层面、不同维度的立法,防控人工智能风险,推动人工智能健康有序、安全可控地发展,以此达到预防人工智能自主性犯罪的目的。AI crime is by no means a“future proposition”.What has caused great controversy in the criminal law community is the issue of AI autonomous crime.In this regard,the“ex post facto punishment”plan believes that the criminal law subject status of AI should be recognized,and on this basis,the criminal punishment system should be reconstructed to deal with AI autonomous crime.The“ex post facto punishment”plan is based on punishing crimes,lacks a realistic basis,and is not desirable.Therefore,AI autonomous crime should be shifted from ex post punishment to ex ante prevention.The core of the“ex ante prevention”plan is risk prevention.Its idea is to prevent and control AI security risks through legislation at different levels and dimensions such as civil law,administrative law,and criminal law,and promote the healthy,orderly,safe and controllable development of AI,so as to achieve the purpose of preventing AI autonomous crime.

关 键 词：人工智能 人工智能自主性犯罪 事后惩罚 事前预防 

分 类 号：D924.3[政治法律—刑法学]