算法合谋的反垄断规制

Antitrust Regulation of Algorithmic Collusion

摘要: 随着人工智能技术逐渐在商业领域的广泛运用，算法对于市场竞争的影响也在日益显著。根据算法在价格协议中的作用不同，可将其分为监控算法、平行算法、信号算法和自主学习算法，由此形成了四种算法合谋类型。这四类算法合谋为执法机关对垄断协议的违法性认定以及归责原则带来挑战。针对算法合谋所造成的反竞争效果和反垄断困境，执法机构应当基于算法的技术原理，对事前监管、事后归责以及违法认定三个方面的体系进行完善或构建，从而为算法合谋的反垄断规制选择一条行之有效的路径。

Abstract: With the gradual and extensive use of artificial intelligence technology in business, the impact of algorithms on market competition is becoming increasingly significant. According to the different roles of algorithms in price agreements, they can be categorized into monitoring algorithms, parallel algorithms, signaling algorithms, and self-learning algorithms, which results in four types of algorithmic collusion. These four types of algorithmic collusion bring challenges for law enforcement agencies to determine the illegality of monopoly agreements and the principle of attribution. In view of the anti-competitive effect and antitrust dilemma caused by algorithmic collusion, law enforcement agencies should improve or construct the system of ex ante supervision, ex post attribution of responsibility, and determination of violation of law based on the technical principle of algorithmic collusion, so as to choose an effective path for the antitrust regulation of algorithmic collusion.