摘要
随着人工智能与大数据技术的高速发展,算法逐渐成为经营者最重要的技术基础和核心竞争力。在显著提高经济效率的同时,算法也为竞争者之间实现价格共谋提供了更加隐蔽的技术手段。高度智能化的自学习算法通过深度学习和机器学习,能在无须人为干预的情况下独立达成共谋,其自主决策的过程如"黑匣子"一般难以被预测和监控。此类新型共谋由算法主导,具有更强的隐蔽性和稳定性,给反垄断领域带来了认定难、监管难、归责难等挑战。通过分析和拓展"协议"的内涵,可明确自主类算法共谋的判定标准,将其纳入反垄断法律的规制范围。并将法律政策与技术手段相结合,构建多元化的规制体系,以促进法律对自主类算法共谋的有效监管。
With the rapid development of artificial intelligence and big data technology,algorithms have gradually become the most important technical foundation and core competitiveness of operators.While significantly improving economic efficiency,algorithms also provide a more covert technical means to realize price collusion among competitors.Through deep learning and machine learning,highly intelligent self-learning algorithms can independently work together without human intervention,and the process of autonomous decision-making,such as the "black box",is generally difficult to predict and monitor.This new type of collusion is dominated by algorithms and has stronger concealment and stability,which brings challenges such as difficulty in identification,supervision and imputation to the antitrust field.By analyzing and expanding the connotation of "agreement",the criterion of autonomous algorithmic collusion can be clarified and the new type of collusion can be incorporated into the scope of anti-monopoly law regulation.Meanwhile,combining legal policy with technical means,and constructing a diversified regulatory system so as to promote the effective supervision of autonomous algorithmic collusion.
作者
沈鸿艺
岳子祺
陈名芮
刘家熳
蔡俊亮
Shen Hongyi;Yue Ziqi;Chen Mingrui;Liu Jiaman;Cai Junliang(South China University of Technology,Guangzhou,Guangdong,510006)
出处
《市场周刊》
2020年第8期164-168,共5页
Market Weekly
基金
教育部国家级大学生创新创业计划项目“自主类算法共谋的反垄断法律规制”(项目编号:201910561033)
华南理工大学百步梯攀登计划项目“对于违法性算法共谋的反垄断法律规制”(项目编号:j2tw201902131)研究成果。
关键词
人工智能
自学习算法
自主类算法共谋
反垄断法
artificial intelligence
self-learning algorithms
autonomous algorithmic collusion
Antitrust Law