Exploring key predictors of ChatGPT adoption in higher education: insights from UTAUT3 and ARCS model

Generative AI is revolutionizing education, and ChatGPT has become widely adopted for learning and academic assistance. Drawing on the Unified Theory of Acceptance and Use of Technology 3 (UTAUT3) framework and the ARCS motivation model, this study examines higher education students' behavioural intention and actual use of ChatGPT. A sequential explanatory mixed-method design was used, integrating quantitative data (based on 455 students from three Pakistani universities) analyzed by using Partial Least Squares—Structural Equation Modelling (PLS-SEM), and qualitative insights generated through semi-structured interviews (n = 41) and thematic analysis. The quantitative results identified effort expectancy, performance expectancy, social influence, hedonic motivation, habit, and ARCS motivation as significant predictors of behavioural intention, while personal innovativeness and facilitating conditions were not significant. ARCS motivation functioned both as a direct determinant and as a moderating factor, amplifying the effects of facilitating conditions, effort expectancy, and social influence on behavioural intention and actual use. The qualitative findings complemented these results by revealing that students valued ChatGPT’s usefulness, ease of use, and peer recommendations but remained cautious regarding accuracy, plagiarism, and unclear institutional policies. Collectively, the synthesized findings illustrate motivation as an essential driver of leveraging external support to achieve sustainable adoption in low-resourced educational settings. By integrating UTAUT3 with the ARCS framework, this study offers a unique holistic perspective on ChatGPT adoption, suggesting that institutional guidance, motivational scaffolding, and ethical considerations are inevitable to optimize AI integration in higher education.