Ethics in the age of generative AI: Student perceptions, use and transparency in South African private higher education

Abstract

Artificial Intelligence (AI) is transforming higher education (HE), raising complex questions about ethics, academic integrity, and responsible use. This study explores students’ perceptions of the ethical use of generative AI (GenAI) tools, focusing on issues of integrity, perceived reliability, quality assurance of outputs, and disclosure practices.  Adopting a survey approach, 1866 students across all disciplines, modes of study and years of study across five PHEIs in South Africa were invited to participate in an online MSForms questionnaire.  Students demonstrated mixed levels of confidence and ethical engagement with GenAI. While ChatGPT was perceived as the most reliable tool (66.5%), many students showed uncertainty regarding its ethical use, with 53.2% remaining neutral about integrity concerns. Gender, study mode, and level of study influenced disclosure practices and perceptions, with males and postgraduate students reporting higher confidence and lower concern. While 66% cross-referenced AI outputs with reliable sources, fewer critically evaluated or proofread them, and 13.5% admitted to non-disclosure.