法学院知识库

迈入数字经济运行的新业态，智能算法已经成为推动经济发展的核心动力之一。算法被广泛应用于商业竞争之中，可以帮助企业提高运行效率，获得市场竞争优势地位。然而，以算法为驱动的合谋行为已经对市场竞争秩序带来了新影响。企业为了获得垄断利润利用算法实施排除、限制竞争的行为，出现多种算法合谋的新类型，难以适用传统的反垄断分析框架，对现有反垄断制度规制带来挑战。

本文以算法合谋概述为基础，梳理了国内外研究现状，阐述算法合谋的内涵、特征，重点界定算法合谋的类型，即信使类合谋、轴辐类合谋、预测代理类合谋、自主学习类合谋。进一步分析我国算法合谋规制现状，明确我国算法合谋反垄断法规制体系，探究算法合谋行为反垄断法规制难点。明确传统的垄断协议概念难以全面规制算法合谋行为，对算法合谋的竞争效果评估和责任主体界定存在争议。同时考察和借鉴美国和欧盟关于算法合谋的相关经验，分析域外算法合谋反垄断法规制原则和执法现状，研究典型案例的规制方法，结合我国反垄断立法和执法情况，对我国算法合谋反垄断法规制提供参考。首先，完善算法合谋垄断协议的法律制度，拓展反垄断法垄断协议的外延，同时重新考量默示合谋的分析方法，进一步建立明确的合法性判断标准；其次，完善利用间接证据判断算法合谋的情形，形成事前收集、事中推定、事后综合考量的全阶段式认定流程；再次，明确算法合谋竞争效果的评估路径，明晰本身违法原则与合理性原则不同类型的评估方法，同时引入消费者福利标准进行竞争效果分析；最后，完善算法合谋垄断责任认定规则，基于算法合谋的类型化分析和场景化界定，明确不同算法合谋主体的责任承担。完善算法合谋的事前规制体系，拓宽数字经济时代新型合谋监管的规制思路。

Stepping into the new business form of digital economy, intelligent algorithm has become one of the core driving forces to promote economic development. Algorithms are widely used in business competition, which can help enterprises improve operation efficiency and gain market competitive advantage. However, the collusion behavior driven by algorithm has brought new influence on the market competition order. In order to gain monopoly profits, enterprises use algorithms to exclude and restrict competition. New types of algorithm collusion appear, which are difficult to apply the traditional antitrust analysis framework, and bring challenges to the existing antitrust regulations.

Based on the overview of algorithm collusion, this paper reviews the research status at home and abroad, expounds the connotation and characteristics of algorithm collusion, and focuses on defining the types of algorithm collusion, namely messenger collusion, axis and spoke collusion, predictive agent collusion and autonomous learning collusion. Further analysis of the current situation of our algorithm collusion regulation, clarify our algorithm collusion anti-monopoly law regulation system, explore the algorithm collusion behavior anti-monopoly law regulation difficulties. It is clear that the traditional concept of monopoly agreement is difficult to fully regulate the algorithm collusion behavior, and there are disputes on the evaluation of the competition effect and the definition of the responsibility subject. Meanwhile, the article examines and learns relevant experience of the United States and the European Union about algorithm collusion, analyzes the regulation principle and law enforcement current situation of foreign algorithm collusion anti-monopoly law, studies the regulation methods of typical case, and provides reference for the regulation of our country's algorithm collusion anti-monopoly law combining the legislation and law enforcement situation in our country. Firstly, we improve the legal system of algorithmic collusion monopoly agreement, expand the extension of antitrust law monopoly agreement, and reconsider the analysis method of implied collusion, and further establish a clear criterion of legality. Secondly, to improve the use of indirect evidence to judge the situation of algorithm collusion, forming a full-stage identification process of collection in advance, presumption in the event and comprehensive consideration afterwards; Thirdly, the evaluation path of algorithm collusion competition effect is clarified, and the different types of evaluation methods of illegal principle and reasonable principle are clarified. At the same time, consumer welfare standard is introduced to analyze the competition effect. Finally, we improve the algorithm collusion monopoly liability identification rules, based on the classification analysis and scene definition of algorithm collusion, to clarify the responsibility of different algorithm collusion subjects. Improve the pre-regulation system of algorithmic collusion and broaden the regulation ideas of new collusive regulation in the era of digital economy.