Exploring student-AI interaction in assessment through the lens of learner agency: case studies in UK higher education

This study explores students-AI interaction, specifically in assessment, through the lens of learner agency, in the context of UK higher education. Using self-regulated learning as the theoretical foundation, this paper used a qualitative approach, namely focus groups, to collect data. Ten focus groups, with forty-two students, were conducted in two Russell Group universities. The findings suggest that even though some have begun moving towards challenging AI outputs and engaging in reflection on the responses provided, the current usage of AI among students appears to be mainly for information retrieval, and/or for accelerating certain routine tasks in their learning. This level of usage therefore only strengthens lower-level thinking skills. The significance of the paper lies in two areas. Firstly, we propose a framework showing the dimensions of student-AI interaction and the connection with self-regulated learning. This framework will influence policy, practice and guidance on AI integration in learning and teaching in higher education. And secondly, this paper serves as guidance on how policy makers and educators can encourage and support effective AI usage among students throughout their learning process.