Artificial intelligence in South African higher education: Survey data of master’s level students

This article presents findings from an online survey conducted at a South African university to explore students' perspectives on the use of artificial intelligence (AI) and generative AI (GenAI) in higher education. A total of 102 (37% of the total cohort) master's students in engineering management provided complete responses. This study builds on the research by Johnson et al. [1], which analyzed 2 555 foundation, undergraduate, structured postgraduate, and research postgraduate students at the University of Liverpool. It enhances their postgraduate findings by specifically examining working professionals in a part-time structured master's program in South Africa. The differing educational and cultural contexts between this study and that of Johnson et al. [1] expand the investigation's scope and facilitate a valuable comparison of student views on AI and GenAI, thus deepening our understanding of how various institutional settings affect technology use in higher education. The findings indicate the students' personal use of AI and GenAI, perspectives on peers' use, and opinions on appropriate policies at the university, program, and module levels. Participants were surveyed on their confidence in academic writing, home languages, and familiarity with AI and GenAI tools. They described how they use these technologies in both personal and academic contexts, the perceived quality of AI-generated content compared to their own work, and their perspectives on others' use of the technologies in an academic setting. Finally, respondents shared their views on the need for institutional regulations to guide ethical and responsible AI usage. Overall, the dataset serves as a valuable resource for education researchers and policymakers seeking to understand students' perceptions and needs regarding AI and GenAI in higher education. Although this survey focused on master's students in engineering management, the questions and findings can be readily adapted to other fields of study and potentially extended to younger student populations, offering insights into support and guidelines necessary to avoid misconduct and overreliance on AI.