算法偏见的法律矫正研究  

Research on Legal Correction of Algorithm Bias

作　　者：赵学刚[1] 马羽男 ZHAO Xuegang;MA Yunan(Law School,Southwest University,Chongqing 400715,China)

机构地区：[1]西南大学法学院,重庆400715

出　　处：《辽宁大学学报（哲学社会科学版）》2020年第5期82-90,共9页Journal of Liaoning University(Philosophy and Social Sciences Edition)

摘　　要：算法技术广泛应用到社会的各个方面,算法的评分、排序决定和影响着相对人贷款、就业、信誉等诸多权益。但当算法做出不利甚至错误决策时,相对人无从知晓原因更遑论改正其错误决策,严重缺乏应有的救济。算法的偏见问题不断地冲击着人们对算法的公共信任程度,于是对于其治理与监管便成为一项重要的法律议题。应创设必要的规制,以遏制算法结果偏见对潜在正当法律利益的威胁和侵害。Algorithm technology is widely used in all aspects of society.The scoring and ranking of algorithms affect many rights and interests of the relative people such as loan,employment,reputation,etc.However,when the algorithm makes unfavorable or even wrong decisions,the relative people have no way of knowing the reason,let alone correcting the wrong decisions,so due relief is seriously lacking.The bias of the algorithm is constantly impacting people’s public trust in it,so its governance and supervision becomes an important legal issue.Therefore,necessary regulations should be created to curb the potential threat and infringement of legitimate interests caused by bias of algorithm result.

关 键 词：算法偏见 自动决策 算法结果拒绝权 大数据 

分 类 号：D925.2[政治法律—诉讼法学]