摘要
人工智能算法促进了共谋的达成和有效实施。从便利共谋实现的角度对算法进行反垄断法意义上的分类,可分为监控式算法、平行式算法、信号式算法和自我学习式算法,在此基础上,形成了信使类共谋、轴辐类共谋、代理类共谋和自主类共谋。对人工智能算法共谋进行反垄断法规制的肇因在于"共谋",而非"算法"本身。对算法共谋进行反垄断法规制时应以"合理规则"为主、"本身违法原则"为辅,对不同类型的算法共谋进行规制时侧重点有所不同。
Algorithms can facilitate the establishment and implementation of collusion.From the perspective of promoting the realization of collusion under anti-monopoly law,algorithms can be classified into monitoring algorithms,parallel algorithms,signaling algorithms and self-learning algorithms which lead to messenger conspiracy,hub-and-spoke conspiracy,the predictable agent collusion and virtual collusion,respectively.The reason for the anti-monopoly regulation of algorithms in the context of artificial intelligence lies in"collusion"rather than"algorithm"itself.The anti-monopoly analysis of algorithmic collusion should be based on"rule of reason"and supplemented by"per se rule".Specifically,the anti-monopoly regulation differs from type to type of algorithms.
作者
刘佳
LIU Jia(Law School,Shandong University of Technology,Zibo 255049,Shandong,China)
出处
《河南大学学报（社会科学版）》
CSSCI
北大核心
2020年第4期80-87,共8页
Journal of Henan University(Social Sciences）
基金
国家社科基金项目“人工智能语境下算法共谋的反垄断问题研究”(19CFX066)阶段性成果
国家留学基金委公派项目“国际经济法视角下互联网产业反垄断问题研究”(CSC No.201808370026)阶段性成果
教育部人文社会科学项目“‘互联网+’语境下滥用市场支配地位的反垄断法规制研究”(17YJC820029)阶段性成果
关键词
人工智能
算法
共谋
反垄断法规制
artificial intelligence
algorithm
collusion
anti-monopoly law regulation
