Academic cheating with generative AI in higher education: An extended model of the theory of planned behavior with motivational antecedents

Abstract

This study proposed and tested an extended model of the theory of planned behavior (TPB) to understand the determinants of academic cheating using generative AI (GenAI). This model integrates intrinsic and extrinsic motivation as antecedents of the core constructs of TPB, namely attitude, subjective norms, and perceived behavioral control, to predict cheating intentions and behavior. Quantitative data were collected from 243 undergraduate students in West Java through a survey and analyzed using confirmatory partial least squares structural equation modeling (PLS-SEM). The model demonstrated satisfactory global fit (SRMR = .045; NFI = .92), supporting the hypothesized structure. The results indicate that the proposed model can explain significant variance in cheating intentions and behavior. Perceived behavioral control proved to be the strongest predictor of cheating intentions. More importantly, both behavioral intention and perceived behavioral control directly and strongly predicted self-reported academic cheating behavior. This study concluded that the extended TPB is a robust framework for this phenomenon, highlighting the dominant role of perceived behavioral control. Its practical implications emphasize the need for institutional interventions focused on reducing the perceived ease and increasing the perceived risk of GenAI misuse to maintain academic integrity.