From Mandatory Exposure to Guided Engagement: Investigating the Structured Integration of Generative AI in Higher Education

Abstract

The rapid institutionalization of generative artificial intelligence (GenAI) in higher education has created an urgent need for empirical evidence on how structured course-level integration relates to student engagement and perceptions of learning. This study examines a usefulness-centered conceptual framework combining elements of the Technology Acceptance Model and the Unified Theory of Acceptance and Use of Technology (TAM/UTAUT) with the digital competence perspective of DigComp 2.2. A structured pedagogical pilot intervention requiring all students to use generative AI tools was implemented in an undergraduate Public Service Management course (n = 76). Students completed AI-supported group assignments and an immediate post-intervention questionnaire comprising 19 Likert-scale items, four demographic questions, and four optional open-ended questions that are not analyzed in the present paper. Because most variables were non-normally distributed, non-parametric statistical methods were applied, including Spearman’s rank correlations, Mann–Whitney U tests, and Kruskal–Wallis tests. Perceived learning usefulness was strongly and positively associated with both frequency of AI use and satisfaction with the learning process. Ethical attitudes were also positive, but more weakly associated with frequency of use. Demographic group differences were observed mainly in specific usage patterns rather than in general attitudes towards AI-supported learning. These exploratory findings suggest that perceived learning usefulness remains relevant in mandatory AI-integration contexts. Pedagogical scaffolding—including prompt literacy, verification practices, and reflective documentation—provides a structured framework for guided and responsible use of generative AI tools in higher education.