Ethical and epistemic AI literacy in student learning judgement, agency, and responsibility in an open distance education context

Abstract The rapid diffusion of generative artificial intelligence (AI) in education has intensified questions regarding how students exercise judgement, responsibility, and agency in AI-supported knowledge practices. While existing research has documented patterns of AI use and perceived usefulness, less attention has been given to how students evaluate AI-generated content, exercise epistemic judgement, and navigate ethical responsibility in learning contexts. This study examines ethical and epistemic AI literacy by analysing how students engage with AI through processes of trust, judgement, and self-regulation, using mathematical learning as an empirical context in which correctness, justification, and verification are central. A mixed-methods descriptive–interpretive design was employed. Data were collected from 341 undergraduates enrolled in a mathematics-related module at a South African open distance e-learning institution. Quantitative survey data was analysed descriptively, and open-ended qualitative responses were examined using reflexive thematic analysis. The findings indicate that students approach AI-generated content with conditional trust, recognising potential inaccuracies while valuing AI as a resource for conceptual understanding. Students report using AI to support learning rather than replace effort, although this orientation is not consistently held. Qualitative accounts suggest that students attempt to regulate their use of AI to maintain responsibility for their work while also expressing uncertainty regarding acceptable practices, particularly in the absence of explicit institutional guidance. The study advances ethical and epistemic understanding of AI use in education by foregrounding learner judgement, agency, and responsibility within an open distance learning context in the Global South. It conceptualises ethical and epistemic AI literacy as an educational competence associated with evaluation, responsibility, and context-sensitive judgement, and it highlights the role of instructional and institutional scaffolding in supporting responsible engagement with AI.