Introduction. The capabilities of modern artificial intelligence (AI)-based solutions to generate texts and solve a range of research tasks have led to their widespread use by students in the preparation of final qualification research papers (thesis) (FQRPs). However, AI-generated solutions for certain research tasks are not always of high quality in terms of content and may violate principles of academic ethics. The aim of this study is to identify the range of research tasks that students currently solve using generative AI tools when preparing their FQRPs. Materials and Methods. The study was conducted using a survey method. An online questionnaire was completed by 782 graduates (2025) from 28 Russian universities across 23 fields of study, including Law, Economics, Linguistics, Pedagogical Education, and others. Respondents were asked to indicate which research tasks they had used AI tools for during their FQRP preparation. KEYWORDS artificial intelligence, qualification paper, ethics, plagiarism Results. The survey yielded the following data: 11.56% used AI tools to generate the full text of their paper; 41.37% for outlining the structure of their paper; 64.21% for sourcing academic references; 17.14% for writing literature reviews; 46.42% for formulating research theses and arguments; 30.15% for developing practical materials (questionnaires, examples, case studies, etc.); 26.15% for statistical analysis of experimental data; 16.31% for data visualization; 56.19% for text editing; 36.66% for formatting bibliographies. Conclusion. The findings highlight the need for regulating AI use in education, including at the institutional level. Legal frameworks should address generative AI tools and define their "legitimate" application in academic and research activities.