Within the last two decades, conflict has become one of the most highly-demanded topics in cross-disciplinary research. Considering the anthropo-centric paradigm as the major research framework of contemporary linguistics, it is apparent that it is necessary to study conflicts as they serve as an indispensable part of society and its systems. Conflict is inherent in interpersonal communication; language is actively involved in the process. While conflict was originally examined in sociology and psychology, it has also attracted linguists, who have explored it across various theoretical frameworks. Political discourse is no exception. On the contrary, political interviews stand out as having significant conflict potential because they represent a kind of dialogical verbal interaction within which a clash of communicative goals appears normal. These interactions employ specialized language tools that serve as markers of conflict communication, utilized by interlocutors to either temper or intensify the conflict, depending on their communicative intentions. The authors propose applying AI-based techniques to identify markers of conflict communications. The authors show the way such techniques may be used for research purposes as far as conflict communication is concerned. The authors singled out such markers of conflict communication as communication styles mismatch, frequency of turn-taking, and different sets of most frequently used words identified by AI-based techniques.