Abstract: The relevance of the study of the mechanisms of interaction of economic agents in the digital space within the limits of behavioral economics is substantiated. Conceptual provisions defining the key parameters of managing the behavior of economic agents in the digital space with the use of artificial intelligence tools within the reflexive characteristics of agents, which determine the peculiarities of their behavior, are proposed. The implementation of relevant conceptual provisions will allow diagnosing the behavior of economic agents and identifying potential agents оf management in the digital space in accordance with the goals of socio-economic systems. The proposed concept is aimed at the use of existing open automated systems of artificial intelligence, if the information is in open or specialized digital databases, and directed questionnaires with appropriate processing of the obtained results using tools of fuzzy logic and neural network modeling, if the information needs to be collected. In turn, management of the behavior of economic agents in the digital space using methods of information and reflexive influence and tools of artificial intelligence will allow to increase the efficiency of achieving the specified goals of socio-economic systems according to the directions of use.

Study of the influence and determination of opportunities and threats of using artificial intelligence tools for analysis and management of the behavior of economic agents in the digital space at macro- and microlevels will reduce the risks of "digital capitalism" and ensure the overall positive socio-economic effect of using the digital space. Prospective directions of research regarding the development of methods and models of the proposed conceptual provisions are outlined.