<p>The study examined the extent to which university students' moral development, religiosity, and AI self-efficacy affect ethical use of AI. Using a cross-sectional research design, data were gathered from 1000 university students, and analysed using partial least squares structural equation modelling (PLS-SEM) and descriptive statistics. The descriptive results revealed that students possess a moderate level of moral development, religiosity, AI self-efficacy and are moderately engaged in ethical use of AI. The PLS SEM analysis, on the other hand, revealed that ethical use of AI is significantly predicted by students' moral development, religiosity, and AI self-efficacy. It was further revealed that AI self-efficacy not only influence ethical use behaviour but also serves as a significant mediating variable that connects religiosity and morality to ethical use behaviours. These results indicate that fostering ethical AI use demands a dual approach: (1) strengthening moral and religious values that underpin academic integrity, and (2) systematically building students’ practical competence and confidence in using AI tools. It is therefore recommended that lecturers should exemplify integrity, incorporate moral case studies and mentor students on ethical decision-making with regard to AI.</p>