Drawing the Line: University Students’ Perspectives on Ethical and Unethical GenAI Use in Academic Work

Abstract

Despite generative artificial intelligence (GenAI) tools being widely used in higher education, ethical boundaries remain unclear. While existing research shows widespread GenAI adoption, less is known about how students define ethical and unethical use, and what criteria they use to distinguish acceptable assistance from authorship substitution. This study examined university students’ perspectives on ethical GenAI use through a thematic analysis of 150 prompt-driven essays collected across three years (2024–2026) from participants at two Croatian universities. Participants demonstrated strong consensus in defining ethical use as functional support for understanding, organization, and technical refinement (brainstorming, clarifying concepts, grammar checking, summarizing materials, structuring essays), and unethical use as content generation that substitutes for student thinking (copying–pasting, exam cheating, submitting AI-generated work). Participants identified three overlapping criteria for authorship substitution: whether AI assists or replaces their work, whether they remain cognitively engaged, and whether they retain intellectual ownership. Participants also expressed concerns about learning loss, cognitive dependence, fairness, and authenticity, while also acknowledging efficiency benefits. Additionally, participants explicitly called for clearer institutional guidance and systematic AI literacy education rather than general prohibitions. Thus, the contribution of this study is the articulation of a student-generated framework for ethical use of GenAI and the translation of that framework into concrete principles for assessment and institutional policy.

https://doi.org/10.26803/MyRes.2026.08