DEVELOPING STUDENTS' INFORMATION SECURITY COMPETENCE IN THE AGE OF GENERATIVE AI: A QUASI-EXPERIMENTAL STUDY IN UZBEK HIGHER EDUCATION

Generative AI tools are now part of students' everyday academic work, but the information security risks involved are rarely covered in curricula. Students leak their own and other people's data into external systems, rely on fabricated output and fall for AI-enhanced social engineering. This paper reports a quasi-experimental study of a pedagogical model designed to develop students' information security competence in the context of AI tools. The model is based on a pedagogical classification of AI related threats into four groups (input data, output, access and account, cognitive-behavioural), a five-stage learning technology in which simulation and reflection are mandatory, and five pedagogical conditions.