Explaining Students' Intention to Use AI-Enhanced E-Learning in Tunisian Higher Education: An Integrative Acceptance Model

Abstract

Purpose. This study explores what shapes Tunisian university students' intention to use e-learning environments that incorporate artificial intelligence (AI). It brings together selected beliefs from the Technology Acceptance Model (TAM) and the Unified Theory of Acceptance and Use of Technology (UTAUT), alongside an exploratory measure of how students perceive AI-related features. Design/methodology/approach. The analysis draws on 200 complete responses from a cross-sectional online convenience survey, as documented in the available analytical summary. Composite-score diagnostics and multiple regression were used, together with a previously reported bootstrap estimate of an indirect association. The outcome is students' stated intention to use these environments, not observed use. Findings. Together, the five predictors explain 57.1% of the observed variance in behavioural intention. The strongest unique association is with AI-related acceptability (β = 0.508), followed by perceived usefulness (β = 0.289) and social influence (β = 0.184). Neither perceived ease of use nor facilitating conditions has a statistically significant unique direct association. The reported indirect estimate linking ease of use to intention through AI-related acceptability is positive (0.277; 95% bootstrap CI [0.184, 0.374]), although the full resampling output is unavailable. Research limitations/implications. The survey relied on a convenience sample, was conducted at one point in time and used a single self-report questionnaire. It therefore cannot support causal claims or population-wide generalisation. Moreover, the AI-related acceptability score combines six different facets, each measured by one item; it should not be treated as a validated multidimensional scale. Privacy concerns were not measured separately, respondents were not asked about one common AI platform, and some diagnostic and mediation details cannot be independently checked. Practical implications. Universities should make the educational benefits of AI-supported learning visible to students, explain how AI features and data practices work, and provide accessible technical and human support. These suggestions are informed by the wider literature rather than tested intervention effects in this survey. Originality/value. The findings suggest that students' overall appraisal of AI-enabled features deserves attention alongside established TAM and UTAUT beliefs. By focusing on Tunisian higher education while acknowledging the limits of its exploratory measure, the article offers a basis for more precise measurement and longitudinal investigation.