Exploring the Adoption and Use of ChatGPT in Higher Education: Factors Influencing Behavioural Intentions and Willingness to Pay Among the Students of Pakistan

The purpose of this study is to investigate the adoption and use of ChatGPT in higher education, with a particular focus on the factors influencing students' behavioural intentions and their willingness to pay for AI-driven educational tools. The research addresses the problem of how individual, psychological, emotional, economic, social, and sociocultural development components shape the acceptance of ChatGPT in academic contexts. These components included 1. General Personal Characteristics, 2. Psychological factors, 3. Perceptions, 4. Emotional intelligence, 5. Career growth and professional development, 6. Intellectual and personal developments, and 7. Sociocultural development. These factors influence core constructs of the UTAUT2 model, which include 1. Performance Expectancy, 2. Effort Expectancy, 3. Social Influence, and 4. Hedonic Motivation. These would, in turn, affect behavioural intention (BI) and use behaviour. We also contend that the effect of these four constructs on behaviour intentions is moderated by interactivity and design in a way that higher levels of both will strengthen the effect of these 4 UTAUT2 constructs on BI. A quantitative approach was employed, targeting undergraduate, graduate and postgraduate students in Pakistan as the primary population, given their active participation in academic learning and exposure to emerging educational technologies. Data was collected using a structured five-point Likert scale questionnaire, and SMART PLS (Partial Least Squares Structural Equation Modelling) was used to test the hypothesised relationships among constructs such as performance expectancy, effort expectancy, social influence, hedonic motivation, perception, and moderating factors including design and interactivity. Significant results indicate that performance expectancy, effort expectancy, social influence, and hedonic motivation directly and positively influence behavioural intention, which in turn strongly predicts actual use behaviour. Among antecedents, general personal characteristics were found to significantly shape effort expectancy and performance expectancy, while psychological and perceptual factors significantly influenced performance expectancy. Emotional intelligence emerged as a robust predictor, positively affecting effort expectancy, hedonic motivation, and social influence, thereby strengthening adoption. Similarly, economic-entrepreneurial considerations were significant only in shaping social influence, highlighting their limited but focused role. In contrast, several hypothesised relationships were statistically insignificant and thus rejected, including the moderating role of design and interactivity, which did not significantly strengthen the linkages between UTAUT2 constructs and behavioural intention. Likewise, most sociocultural and social growth components showed no significant effect on expectancy dimensions or hedonic motivation. Overall, the study demonstrates that students' adoption of ChatGPT is primarily driven by usefulness (PE), ease of use (EE), enjoyment (HM), and social influence (SI), reinforced by emotional intelligence and personal characteristics, while contextual moderators like design and interactivity, as well as broader sociocultural factors, did not play a decisive role. This research contributes to the literature on technology adoption in higher education by providing empirical evidence from Pakistan and offering insights for educators, policymakers, and developers seeking to integrate AI tools into learning environments.