AI Literacy and Scenario-based Ethical Judgment About Generative AI Among Undergraduate Students

Abstract

Generative artificial intelligence (GenAI) is increasingly embedded in higher education, shaping how students learn and obtain knowledge. Although AI education is expanding, limited empirical research has examined how undergraduate students exercise ethical judgment about using GenAI in academic situations, particularly in relation to their AI literacy. This study examined undergraduate students’ ethical judgment regarding GenAI use in educational contexts and its relationship to AI literacy and student demographics. A cross-sectional survey was conducted with a convenience sample (n = 531) recruited through two psychology courses at one public Midwestern university. The survey included demographics, six GenAI ethical scenarios, and the Generative AI Literacy Assessment Test (GLAT). Participants generally demonstrated high agreement when classifying the six predesignated scenarios, distinguishing GenAI use that supported learning from use that substituted AI-generated content for original academic work. 57.6% of students correctly classified all six scenarios, indicating a pronounced ceiling effect, with uncertainty greatest when AI-generated content was minimally modified before submission. AI literacy demonstrated a small positive association with classification accuracy, Spearman’s ρ = 0.234, p < .001, and emerged as a significant predictor across the six scenarios. In the adjusted binomial model, each additional correct response on the objective AI literacy measure was associated with a 13.9% increase in the modeled odds of correctly classifying an ethical scenario. By focusing on applied ethical judgment in academic scenarios, this study highlights the need to prospectively evaluate how AI literacy instruction and ethics education may support undergraduate students’ preparation for future professional practice.