Undergraduate students’ perspectives of generative AI ethics

Abstract The prevalence of ChatGPT uses in higher education prompted instructors to apply AI ethics to their classroom, which are moral standards of AI uses. Yet, as a primary group of AI users, undergraduates’ perspectives have rarely been explored to help instructors integrate AI ethics into the classroom. Using a mixed-method approach, this study explored 110 undergraduate students’ agreement levels on five AI ethics from a previous study, the factors influencing their agreement levels, and the rationales explaining their choices. The study showed that undergraduate students agreed with “autonomy” most and “justice” least; gender and grade level could predict their agreement levels; moreover, 14 themes were discovered regarding students’ rationales. Its results enlighten instructors to highlight public interests, consider students’ diversity, develop specific guidelines, and facilitate attitudinal changes in the process of applying instructions about AI ethics.