Higher education has undergone significant transformation as a result of the recent development of generative artificial intelligence (GenAI) tools such as ChatGPT. In the literature, the pedagogical and ethical effects of incorporating GenAI into educational assessments are still not well understood. This is the gap in literature that this study intends to contribute to. The AI Assessment Scale was developed with an aim to assist lecturers in helping students utilise GenAI tools in a responsible way, taking in consideration academic integrity and ethics.

The purpose of the study was to examine generative AI use in assessments. The interpretivism paradigm was applied because it focuses on how individuals construct their knowledge through personal experiences and reality is socially constructed. Qualitative case study design was selected for this study because it is associated with the interpretivist perspective. The qualitative approach is a useful tool when examining and understanding the meaning that individuals or groups attribute to social or human problems. The case study is in-depth analyses of individuals or groups of individuals that supports researchers in understanding the background, methods, and findings of a study. The participants in this study comprised 28 students in the Faculty of Science at a University of Technology in South Africa. These participants were selected using purposive and convenience sampling. Purposive sampling was applied because the researcher selected only students who were deemed fit and knowledgeable about the subject being investigated. Furthermore, the researcher selected these participants conveniently because they were easily accessible where the researcher worked. A survey questionnaire with open-ended questions was used to gather and compare insight from the participating students in 2023 and 2024. Thematic analysis using Atlas.ti was applied to compare and analyse data. It was found from both groups of students that the text-generative AI ChatGPT was the most used AI tool. The findings revealed that students in both groups used AI to get more information about the course or the concepts that they were engaging with, and it assisted when writing assignments. On the contrary, the finding showed that half of the students did not use AI for learning. To ensure that students integrate AI responsibly and effectively and to enhance their academic performance, higher education institutions should adopt policies, ethical usage, and guidelines for using AI in education.

Keywords: Generative AI, Assessment, education 6.0.