Motivational and normative drivers of generative AI substitution in academic work: a mixed-methods study from Saudi higher education
TL;DR
This mixed-methods study investigates how attitudes toward AI, moral rationalization strategies, and perceived institutional clarity interact to shape AI-based substitution behavior among 249 undergraduates at a Saudi university to offer a new approach to evaluating the effectiveness of institutional AI guidance beyond common technology acceptance models.
Abstract
The rapid integration of generative artificial intelligence (GenAI) in higher education has intensified tensions between legitimate learning support and unauthorized task substitution, particularly where institutional guidance remains ambiguous. This mixed-methods study investigates how attitudes toward AI, moral rationalization strategies, and perceived institutional clarity interact to shape AI-based substitution behavior among 249 undergraduates at a Saudi university. Partial least squares structural equation modeling (PLS-SEM) revealed that positive attitudes and rationalization together explained 46% of the variance in substitution behavior, with perceived clarity of institutional guidance significantly moderating the rationalization–substitution link. Complementary interviews with seven students and ten instructors revealed that linguistic burden, peer norms, and fragmented faculty guidance facilitated boundary crossing from scaffolding to shortcutting. By exploring the relationship between moral neutralization and environmental clarity, this research offers a new approach to evaluating the effectiveness of institutional AI guidance beyond common technology acceptance models. The findings can be used to inform the design of multi-tiered, inclusive AI usage policies, assessments that value process over product, and culturally responsive academic integrity education within a multilingual higher education context.