Generative AI in Tunisian Higher Education
Part One: Between Educational Transformation and Ethical Issues
DOI:
https://doi.org/10.26034/fr.jehe.2026.9709Keywords:
ethics in higher education, Gen AI, Generative AI technology, Tunisian higher education, Partial Least Squares Structural Equation Modeling, PLS-SEMAbstract
This study analyzes the determinants of AI adoption among 221 Tunisian students using an integrated model combining TAM, UTAUT, TOE, as well as ethical dimensions and perceived risks. Through a PLS-SEM approach, it shows that behavioral intention strongly influences actual use (β = 0.588), while facilitating conditions also exert a significant effect (β = 0.249). In contrast to classical models, perceived usefulness and perceived ease of use are not significant. Conversely, perceived ethics (β = 0.244), habit (β = 0.214), and the moderating effect of organizational support (β = 0.189) emerge as key determinants. The inclusion of control variables confirms the robustness of the model and reveals contextual effects, particularly the influence of environmental pressure, social influence, gender, and field of study. The model demonstrates high explanatory power (R²BI = 0.551; R²UR = 0.522) and highlights a shift in determinants toward ethical, behavioral, and contextual dimensions.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 Mohamed Ali Bahrini, Eya Soltani
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
Journal articles of Globethics Publications are published under the open Creative Commons License Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0), which guarantees the rights of licensor and allows free use and re-use to the licensees (the readers) who can: 1) Share — copy and redistribute the material in any medium or format 2) Adapt — remix, transform, and build upon the material provided appropriate credit is given and similar license is used in case of such adaptations. Content should not be used for commercial purposes. Each article (the version of record) can be deposited by the author on their academic institutional repository or personal author webpage.
