Algorithmic Collusion: The Determination of Boundary and Its Regulation of Antitrust Laws
-
摘要: 市场主体通过算法利用市场透明度使共谋不受市场集中度的限制, 突破传统共谋的形式而达成新型算法共谋, 破坏市场竞争秩序, 因此有必要通过反垄断法对算法共谋实施规制。在反垄断法规制算法共谋边界的确立上, 可以将算法共谋类型化为信使类共谋、轴辐类共谋、预测类共谋、自主类共谋, 探寻其中真实主观意图以及是否存在垄断协议来准确认定算法共谋。在具体实施方案的选择上, 反垄断执法机关应秉持谦抑执法理念, 将算法实施者作为算法共谋的责任主体, 通过有限公开算法、采用监管科技等具体措施, 对算法共谋实施合理的反垄断法规制。Abstract: Market subjects make use of market transparency and algorithm to make collusion not restricted by market concentration, which breaks through the traditional form of collusion to achieve new algorithm collusion and destroy the order of market competition, so it is necessary to regulate algorithm collusion through anti-monopoly law. On the establishment of the collusion boundary of the anti-monopoly law, the collusion types of the algorithm can be classified into messenger collusion, axial and radial collusion, prediction collusion and independent collusion, and the real subjective intention and the existence of monopoly agreement can be explored to accurately identify the collusion of the algorithm. In terms of the selection of specific implementation plan, the anti-monopoly law enforcement agencies should adhere to the principle of modest law enforcement, take the algorithm implementer as the main body of responsibility for algorithm collusion, and implement reasonable anti-monopoly regulations on algorithm collusion through specific measures such as limited disclosure of algorithm and regulatory technology.
-
Key words:
- algorithm /
- algorithmic collusion /
- algorithm regulation /
- modesty /
- anti-monopoly law /
- boundary
-
表 1 域外算法共谋案件一览表
编号 时间 国别/地区 案件 1 2015 美国 U.S. v. Daniel William Aston and Trod Limited 2 2015 美国 Spencer Meyer v. Uber 3 2015 美国 AirLine Tariff Publishing 4 2016 美国 U.S. v. Topkins 5 2018 欧盟 ECJ’s Judgment in Case C-74/14, Eturas 6 2018 欧盟 Case AT.40465-ASUS 7 2018 欧盟 Case AT.40182-Pioneer 8 2018 欧盟 Case AT.40181-Philips 9 2018 欧盟 Case AT.40469-DENON & MARANTZ 10 2019 美国 U.S. v. G Nova Corporation 11 2019 美国 U.S. v. Thales S.A. and Gemalto N.V. 12 2019 美国 National Fair Housing Alliance v. Facebook, Inc. (S.D.N.Y.) 
下载: 导出CSV
表 2 算法共谋形式
个, % 共谋形式 数量 百分比 有效百分比 累计百分比 信使类共谋 7 58.3 58.3 58.3 轴辐类共谋 4 33.3 33.3 91.7 预测类共谋 1 8.3 8.3 100.0 自主类共谋 0 0 0 100.0 总计 12 100.0 100.0
下载: 导出CSV
-
[1] ROBERT SEDGEWICK, KEVIN WAYNE.算法[M].第4版.谢路云, 译.北京: 人民邮电出版社, 2012. [2] 沈伟伟.算法透明原则的迷思---算法规制理论的批判[J].环球法律评论, 2019(6): 20-39. https://www.cnki.com.cn/Article/CJFDTOTAL-WGFY201906003.htm [3] 陈鹏.算法的权力:应用与规制[J].浙江社会科学, 2019(4): 52-58, 157. https://www.cnki.com.cn/Article/CJFDTOTAL-ZJSH201904007.htm [4] ARIEL EZRACHI, MAURICE E STUCKE. Two Artificial neural networks meet in an online hub and change the future(of compe-tition, market dynamics and society)[J]. University of tennessee college of law, 2017, 7(323): 1-15. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2949434 [5] NICOLA PETIT. Antitrust and Artificial intelligence: a research agenda[J]. Journal of european competition law and practice, 2017, 8(6): 361. doi: 10.1093/jeclap/lpx033 [6] AZADEH NEMATZADEH, GIOVANNI LUCA CIAMPAGLIA, FILIPPO MENCZER. How algorithmic popularity bias hinders or promotes quality[J]. Scientific reports, 2017, 8(1): 21. http://www.ncbi.nlm.nih.gov/pubmed/30374134 [7] 黄晓伟.互联网平台垄断问题的算法共谋根源及协同治理思路[J].中国科技论坛, 2019(9): 9-12. https://www.cnki.com.cn/Article/CJFDTOTAL-ZGKT201909005.htm [8] 韩伟.算法合谋反垄断初探---OECD《算法与合谋》报告介评(上)[J].竞争政策研究, 2017(5): 114-123. https://www.cnki.com.cn/Article/CJFDTOTAL-JZZC201705034.htm [9] 李婕.垄断抑或公开:算法规制的法经济学分析[J].理论视野, 2019(1): 66-69. https://www.cnki.com.cn/Article/CJFDTOTAL-LLSY201901012.htm [10] 李振利, 李毅.论算法共谋的反垄断规制路径[J].学术交流, 2018(7): 73-82. doi: 10.3969/j.issn.1000-8284.2018.07.012 [11] 施春风.定价算法在网络交易中的反垄断法律规制[J].河北法学, 2018(11): 113-121. https://www.cnki.com.cn/Article/CJFDTOTAL-HBFX201811010.htm [12] 朱理.网络经济背景下的反垄断司法[J].财经问题研究, 2018(9): 5-8. [13] 张蕴萍.规制能力提升是深化中国垄断行业政府规制体制改革的有效途径[J].理论学刊, 2015(8): 43-48. doi: 10.3969/j.issn.1002-3909.2015.08.006 [14] DENG A. What do we know about algorithmic tacit collusion?[J]. Social science electronic publishing, 2018, 33(1): 55-65. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=3171315 [15] 韩伟.数字市场竞争政策研究[M].北京: 法律出版社, 2017: 78. [16] 韩伟.数字经济时代中国《反垄断法》的修订与完善[J].竞争政策研究, 2018(4): 53-64. https://www.cnki.com.cn/Article/CJFDTOTAL-JZZC201804005.htm [17] CALVANO E, CALZOLARI G, DENICOLO V, et al. Algorithmic pricing what implications for competition policy?[J]. Review of industrial crganization, 2019, 55(1): 155-171. doi: 10.1007/s11151-019-09689-3 [18] 韩伟.算法合谋反垄断初探---OECD《算法与合谋》报告介评(下)[J].竞争政策研究, 2017(6): 68-77. https://www.cnki.com.cn/Article/CJFDTOTAL-JZZC201706012.htm [19] 叶卫平.反垄断法分析模式的中国选择[J].中国社会科学, 2017(3): 97-116, 207. https://www.cnki.com.cn/Article/CJFDTOTAL-ZSHK201703005.htm [20] 罗伯特·诺奇克.合理性的本质[M].葛四友, 陈昉, 译.上海: 上海译文出版社, 2012: 6. [21] 曾雄.人工智能时代下算法共谋的反垄断法规制[J].网络法律评论, 2016(2): 16-27. https://www.cnki.com.cn/Article/CJFDTOTAL-WLFP201602006.htm [22] FLEISCHACKER S. On Adam Smith's Wealth of Nations: a philosophical companion[M]. New Jersey: Princeton University Press, 2009: 157. [23] KAPLOW L. On the meaning of horizontal agreements in competition law[J]. Social science electronic publishing, 2011, 99(3): 683-818. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1873430&rec=1&srcabs=1873412&pos=1 [24] 丁茂中.反垄断法实施中的"积极失误"与"消极失误"比较研究[J].法学评论, 2017(3): 80-86. https://www.cnki.com.cn/Article/CJFDTOTAL-FXPL201703008.htm [25] 邓志松, 戴健民.数字经济的垄断与竞争:兼评欧盟谷歌反垄断案[J].中国市场监管研究, 2017(10): 33-36. https://www.cnki.com.cn/Article/CJFDTOTAL-GOGY201710005.htm [26] 孙晋.谦抑理念下互联网服务行业经营者集中救济调适[J].中国法学, 2018(6): 151-171. https://www.cnki.com.cn/Article/CJFDTOTAL-ZGFX201806009.htm [27] 杨立新.人工类人格:智能机器人的民法地位---兼论智能机器人致人损害的民事责任[J].求是学刊, 2018(4): 89-101. https://www.cnki.com.cn/Article/CJFDTOTAL-QSXK201804011.htm [28] 赵万一.机器人的法律主体地位辨析---兼谈对机器人进行法律规制的基本要求[J].贵州民族大学学报:哲学社会科学版, 2018(3): 152-172. https://www.cnki.com.cn/Article/CJFDTOTAL-GZMZ201803011.htm [29] 吴汉东.人工智能时代的制度安排与法律规制[J].法律科学(西北政法大学学报), 2017(5): 130-138. https://www.cnki.com.cn/Article/CJFDTOTAL-DOUB201705012.htm [30] METZGER M J, WILSON C, ZHAO B Y. Benefits of browsing?the prevalence, nature, and effects of profile consumption behavior in social network sites[J]. Journal of computer-mediated communication, 2018, 23(2): 72-89. doi: 10.1093/jcmc/zmx004 -
