Artificial intelligence in science education: a study of opportunities and challenges in a South African University.

Abstract

The integration of Artificial Intelligence (AI) into education is reshaping teaching and learning worldwide, offering innovative solutions to enhance academic outcomes. This study explores the perceptions of undergraduate science students at the University of KwaZulu-Natal regarding the use of AI in science education, with a specific focus on its opportunities, challenges, and ethical and social implications. Guided by Social Constructivism and Machine Learning Theory, the study aimed to investigate the opportunities and challenges of incorporating AI into science education, how AI applications can personalise and enhance learning experiences, and the ethical and social considerations involved in AI use. A mixed method approach was employed, combining quantitative data collected through structured questionnaires and qualitative data from semi-structured interviews. Participants were selected using purposive and convenience sampling. The results revealed strong student support for AI’s potential to personalise learning, improve understanding of complex science content, and increase engagement. However, several challenges were identified, including limited access to digital infrastructure, concerns about over-reliance on AI, academic dishonesty, and data privacy risks. Based on the findings, the study recommends improving institutional infrastructure, providing digital literacy and ethical AI use training, and integrating AI tools into science curricula in a balanced and supportive manner. These recommendations aim to ensure equitable access, promote critical thinking, and support responsible AI use in higher education. The study contributes to the growing body of research on AI in education by offering context-specific insights that can inform policy, curriculum development, and future research.