Аннотация:In recent years social bots became one of the most popular political manipulation methods to change public opinion on the Internet. Traces of social bots’ usage were found not only in authoritarian but also in democratic countries, and these accounts, by imitating ordinary people, were able to spread disinformation and suppress opposition agenda all around the world. Despite the prevalence of publications and articles in various newspapers and scientific journals focusing on the results of the social bots’ interference, seldom reports explain the process by which social bots, while being outcasts at first, are able to spread fake messages all over the network. Therefore, the aim of this study is to examine the factors which contribute to the process of social bots’ intervention and radicalization of the Net by creating and analyzing an agent-based model of political communication. Bounded rationality, the spiral of silence, and echo chambers concepts are used as a theoretical foundation of our model. The results suggest that Social bots can significantly change the users’ political position. It requires at least 10% of political bots from one side and no bots from the other to completely radicalize the network over long periods of time. There are 2 main polarization scenarios: (1) even radicalization of the social network with a layer of neutral ordinary users between polarized groups and (2)complete radicalization of the Network with oppositional leaders being ostracized. As for the role of initial connectivity between users, with a greater number of users’ connections, social bots’ infection rate becomes slower but they form more stable networks of supporters when the final quantity is less dependent on random factors