Beyond AI Use: Understanding University Students’ Acceptance of AI-Assisted Academic Misuse

Abstract

The rapid adoption of generative artificial intelligence (AI) in higher education has created new learning opportunities while raising concerns about academic integrity and students’ acceptance of AI-assisted academic misuse. This study examined factors associated with such acceptance among 304 university students in Kuwait using a cross-sectional online survey and non-probability convenience and snowball sampling. The three focal constructs—general academic-integrity awareness, policy and AI-plagiarism awareness, and acceptance of AI-assisted academic misuse—were measured using 15 dichotomous Yes/No items coded 0 = No and 1 = Yes. Acceptance of AI-assisted academic misuse was operationalised as the mean of three misuse-acceptance items (Y12-Y14), producing a composite ranging from 0 to 1 that was analysed as a continuous outcome. Data were analysed using descriptive statistics, principal-axis exploratory factor analysis with varimax rotation, KR-20 reliability analysis, correlations, and ordinary least squares regression with HC3 robust standard errors. The regression model was significant (F = 2.579, p = .010, R² = .067). Policy and AI-plagiarism awareness was positively associated with misuse acceptance (B = .193, p = .013), contrary to the hypothesised negative direction, and Kuwaiti nationality was also associated with higher acceptance (B = .103, p = .039). General academic-integrity awareness and prior AI use were not significant. Because the study is cross-sectional and the misuse-acceptance scale had modest reliability (KR-20 = .587), the findings are interpreted as associations rather than causal effects. The results support clearer, context-specific AI policies and guidance that distinguishes legitimate AI-supported learning from academically inappropriate use