On Anti-monopoly Law Regulation Of Algorithmic Collusion

In the past decade,the development of information technology can be said to be rapid.With the continuous improvement of network services and the continuous growth of data volume,artificial intelligence algorithms with inherent advantages in data processing emerge as the times require and become increasingly powerful.After adopting the algorithm,the enterprise improves the automation of its own production,provides products more in line with the market demand,and satisfies the needs of consumers in many aspects.However,in addition to the positive impact,the use of algorithm also brings hidden risks to the good market order.One of the risks is that enterprises collude with each other to implement monopoly behavior.Algorithm is widely used in various industries,which greatly increases the possibility of collusion between enterprises,and collusion has always been the key regulatory object of anti-monopoly law.The regulation of new collusion implemented by algorithm is the core of this paper.At the present stage,China’s anti-monopoly law is slightly inadequate in regulating collusion,and is facing many difficulties.Based on the traditional collusion theory,economic theory and computer science and technology,this paper makes an in-depth analysis and Research on algorithmic collusion,and puts forward reasonable regulatory suggestions combined with China’s anti-monopoly practice.The first part is the research on the concept、types and market harmfulness of algorithm collusion.By discussing the influence of algorithm in the market and the anti-competitive application of enterprises,the connotation of algorithm collusion and four kinds of algorithm types affecting collusion are defined.Finally,it proposes the way to strengthen the market structure by combining the tacit agreement theory with the traditional theory.The second part mainly analyzes the dilemma of China’s anti-monopoly law in regulating pricing algorithm collusion.The first is the limitation of monopoly agreement system in China.On the one hand,the concept of "agreement" in China’s existing monopoly agreement system can not completely cover all the cases of algorithm collusion,and because China’s monopoly agreement system adopts the dichotomy,it is unable to accurately identify its nature in the face of third-party led pricing algorithm collusion.Secondly,the traditional monopoly law enforcement requires the identification of the subjective monopoly intention of enterprises,which becomes difficult in some types of collusion.Finally,the existing analysis tools of self violation and reasonable principle are difficult to apply when dealing with algorithm collusion.The third part puts forward the corresponding suggestions from two aspects:the principle of deterministic analysis,the subjective identification strategy and the construction of the standard system for the algorithm itself.The first is to change the traditional way of judging whether collusion is based on the illegal act of agreement,but to investigate whether the algorithm fully fulfills the relevant system requirements based on the type of behavior in the case.In terms of principle application,only "messenger" collusion adopts the principle of self violation,and the other three adopt the principle of rationality with different degrees of strictness.Secondly,it puts forward the obligation of explanation for the enterprise application algorithm,and establishes the registration and compliance review system for the algorithm,so as to reduce the collusion risk of the algorithm from the source.