Investigating Students' Intentions to Use Generative AI in College Assignments: A Moderated Mediation Analysis of Perceived Unfairness and Detection Possibility

Objectives Drawn from the theory of planned behavior (TPB), this study aims to examine the students’ attitudes and behaviors regarding the acceptance and use of generative artificial intelligence (AI) in college assignments. Methods A survey with 193 students from a South Korean university was conducted with both qualitative and quantitative methods. In terms of the qualitative method, a thematic analysis was conducted. For quantitative method, a moderated mediation model with path analysis was tested using Mplus software. Results The quantitative analysis revealed that students perceiving high unfairness in generative AI and those anticipating a high possibility of detection in generative AI assignments were more inclined to express lower acceptance of generative AI. Furthermore, the relationship between perceived unfairness and AI acceptance was moderated by the perception of detection possibility. A heightened possibility of detection strengthened the negative association between perceived unfairness and AI acceptance. The study also found that the acceptance of generative AI significantly increase the actual behavior of employing this technology for assignments. The qualitative analysis of open-ended questions demonstrated that students generally exhibited positive receiving attitudes to generative AI, but they also recognized potential side effects and acknowledged the need for academic regulations. Additionally, students noted shortcomings in the credibility and practicality of current generative AI technology. Conclusions This research is one of the first attempts that tackled students' motivation towards AI and provides the educational implications for use of AI in pedagogical purposes.