The Risks and Responses of Algorithmic Discrimination from the Perspective of Criminal Regulation
The application of artificial intelligence(AI)has a positive effect on improving judicial efficiency and em-powering judicial productivity,but its internal algorithms may be embedded with cognitive bias,improper program design,and biased function settings,which will lead to algorithmic discrimination,and thus impact on the traditional criminal law system and the theory of procedural justice.This is specifically manifested in the two aspects of legal entities and legal pro-cedures.At the entity level,traditional criminal behavior is easy to be alienated,the protectionof legal interests is easy to be out of order,and the identification of the subject of criminal responsibility is easy to be in trouble.At the procedural lev-el,the openness and neutrality of adjudication are easily eroded,and the confrontation between the prosecution and the de-fense is easily imbalanced,thus weakening the mobility and independence of the judiciary.Therefore,the application of ar-tificial intelligence should be strictly criminal regulation,the focus of regulation should be placed on the human behavior be-hind the algorithmic technology.Within the framework of the existing criminal law doctrine,the dual regulation of the risk of algorithmic discrimination should be constructed both ex ante and ex post by activating the explanatory tension of criminal law offenses,reinforcing the criminal judicial review of algorithmic discrimination,and granting new types of algorithmic rights to algorithmic counterparts.
artificial intelligencealgorithmic discriminationcriminal law theoryprocedural justice
NETL
NSTL
万方数据
