Mapping Artificial Intelligence Research in STEM Education in South Africa

Abstract

Artificial Intelligence (AI) is increasingly influencing education systems around the world, with growing applications in teaching, learning, assessment, and school management. In South Africa, there is a rising interest in how AI can be used in education, especially in response to national goals tied to digital transformation. Yet, how South African researchers are engaging with AI in education, particularly in the context of Science, Technology, Engineering and Mathematics (STEM) education, remains insufficiently understood. This study uses a systematic bibliometric approach to explore the research on how AI has been applied in STEM education within South Africa. Based on analysis of Scopus dataset, we used VOSviewer and R Studio to examine publication trends, research collaborations, thematic priorities, and emerging areas of focus. The findings show a notable increase in research output after 2020.  Most of the high-output authors are affiliated with a small number of South African universities, and collaboration across institutions remains limited. By analysing how keywords are used across the literature, the study identified several common themes, such as digital transformation, machine learning, and issues around ethics. The findings further suggest that AI application research remains limited in inclusive education, curriculum development and classroom practices. Interdisciplinary collaborations in STEM subjects at elementary school level within the South African context must be strengthened.  These insights can help guide educators, researchers, and policymakers to ensure that AI tools not only align with South Africa’s broader educational goals but also contribute in ways that are practical, inclusive, and sustainable.