Exploring the Use of Generative Artificial Intelligence (GenAI) in Teaching, Learning and Assessment of STEM Subjects

Abstract

The rapid rise of generative AI (GenAI) tools presents both opportunities and challenges for transforming the teaching, learning and assessment (TLA) of STEM subjects. This mixed-methods study examined the use of GenAI at Samtse College of Education (SCE), Bhutan, drawing on survey responses from 147 STEM students (ICT: n = 96; Science: n = 51) and four focus group interviews. The study investigated the integration, purposes, comfort and frequency of GenAI use, as well as the associated impacts, challenges and limitations. Findings indicate that SCE STEM students are rapidly integrating GenAI into their academic practices, with ChatGPT serving as the primary tool for assignment support, academic writing and information access. The results highlight the versatility and perceived usefulness of GenAI, while also pointing to risks such as overdependence, reduced tutor-student interaction and ethical concerns. Subject discipline and academic level, rather than gender, emerged as the strongest predictors of comfort and frequency of use. The study recommends establishing clear policies on academic integrity, acceptable use of GenAI, and data privacy and security, while providing students and faculty with clear guidelines to navigate both opportunities and risks.