GENERATIVE ARTIFICIAL INTELLIGENCE IN FINAL QUALIFICATION WORKS: POTENTIAL, RATIONAL USE, LEGAL REGULATION

The relevance of the research is related to the rapid growth of AI-tools availability and the expansion of the number of their users among students, as well as to the improvement of the quality of text generation combined with the absence of a state position and social consensus on the key issues of AI use in the educational process, primarily in the course of the state final assessment. The main goal of the research is to identify the educational potential of generative AI in the preparation of final qualification works, to reveal the content of the concept of «rational use» and to formulate the criteria of rational use, as well as to substantiate the most practical approach to the legal regulation of the emerging relations, based on the results of the analysis of normative acts, statistical data and doctrinal sources. The problems under consideration: the duality of the potential of generative AI in educational and research activities; the criteria for the rational use of generative models in writing final qualification works; the correlation of the established approaches to regulation and the prospects of university local rule-making. The research is based on the dialectical method of cognition, using general scientific and special methods. Conclusions. Rational use of generative AI in the preparation of final qualification works presupposes compliance with the requirements of controllability, limitation and good faith (the COD-test). The global trend is the gradual transition of university rulemaking from the prohibitive-controlling to the integrative-learning approach; in the absence of legislative regulation, local acts of universities remain the main regulator of the relations under consideration. The proposed COD-test, which enshrines the requirements of controllability, limitedness and good faith in the use of generative models, may serve as the substantive foundation for developing local acts of educational organizations, while the systematized description of the research and educational potential of generative AI may underpin further scholarly discussions in the development and debate of the corresponding ethical concepts