Academic Misconduct and Generative Artificial Intelligence: University Students’ Intentions, Usage, and Perceptions [Preprint]

This study aimed to investigate the relationship between academic misconduct and the use of Generative Artificial Intelligence (Gen-AI) among university students. In the current study, we tested the hypotheses that students with pre-existing misconduct intentions and behaviors were: (1) more inclined to use Gen-AI for assessments; (2) more likely to have already used Gen-AI for assessments; and (3) less likely to view Gen-AI as cheating. Accordingly, we surveyed 442 undergraduates using two subscales of the academic misconduct scale to assess intentions and behaviors related to academic misconduct. Questions also addressed their intent and current use of Gen-AI for university assessments and perceptions of Gen-AI as cheating. Findings showed that students with past misconduct intentions or behaviors were 270% and 138% more likely, respectively, to consider using Gen-AI. However, past misconduct behaviors showed a non-significant 66% increase in current Gen-AI usage. Moreover, students with misconduct intentions were 50% less likely to view Gen-AI as cheating, while prior misconduct showed a non-significant 38% increase in this perception. The results emphasize the need for universities to guide students towards ethical Gen-AI use, especially considering it is being used by those predisposed to academic dishonesty.