Purpose The rapid integration of generative artificial intelligence (AI) tools, such as ChatGPT, into higher education presents both new opportunities and ethical challenges. This study aims to examine university students’ perceptions and ethical readiness toward the use of generative AI tools in the educational context, with a comparative focus on Science, Technology, Engineering, and Mathematics (STEM) and non-STEM disciplines. Design/methodology/approach The research uses the Ethical Adoption Model of Generative AI, an extension of the unified theory of acceptance and use of technology 2 (UTAUT2) framework that integrates ethical constructs, including ethical awareness (EA), perceived ethical risk (PER) and AI ethical anxiety (AIEA). Data were collected from 372 university students in Indonesia. Quantitative analyses, comprising t-tests, analysis of variance and partial least squares structural equation modeling (PLS-SEM), were used to examine both group differences and predictive relationships among constructs. Findings Results revealed significant differences between STEM and non-STEM students across several dimensions, particularly hedonic motivation, facilitating condition, social influence and AI ethical anxiety. Habit emerged as the most distinctive factor among STEM students, indicating deeper integration of AI into their academic routines. The PLS-SEM results identified performance expectancy, ethical awareness, habit and AI ethical anxiety as significant predictors of behavioral intention, with AI ethical anxiety exerting a negative influence. The findings underscore that STEM and non-STEM students experience distinct ethical concerns and adoption patterns, suggesting the need for differentiated institutional strategies, emphasizing technical empowerment for STEM students and ethical guidance for non-STEM students, to promote responsible and equitable AI adoption in higher education. Originality/value This study integrates UTAUT2 and ethical decision-making constructs to reveal an ethical mediation chain (Ethical Awareness → Perceived Ethical Risk → AI ethical anxiety) shaping students’ behavioral intention.
Qomariyah et al. (Fri,) studied this question.