The integration of artificial intelligence (AI) into assessment practices is rapidly reshaping higher education, raising important questions about the nature of learning, feedback, and academic agency. While AI-enabled assessment is often positioned as a tool to enhance efficiency and support learning, its implications for self-directed learning (SDL) remain underexplored. This qualitative study investigates an AI-enabled assessment in a first-year distance teacher education module in which AI was used both to generate and evaluate student work. Drawing on data from 128 student reflections collected as part of the assessment, and analysed using a reflexive thematic analysis approach, the study examines how students engaged with AI within the assessment process. The findings indicate that AI can support key processes associated with self-directed learning, including reflection, self-monitoring, and feedback engagement, but that this support is often structurally mediated rather than internally driven, revealing a critical tension. Students demonstrated both agency and uncertainty, highlighting the complexity of learning in AI-mediated environments. To interpret these findings, the study proposes the FRAME-AI model (Facilitator-Regulated Assessment Model for AI-enabled learning), which conceptualises AI-enabled assessment as a pedagogically mediated process shaped by intentional design, transparent communication, guided AI engagement, feedback mediation, and facilitator oversight. The model positions the facilitator as central in shaping how AI is experienced and used within the learning process. The study contributes to emerging scholarship by challenging the assumption of AI as a neutral educational tool and by offering a conceptually grounded framework for designing AI-enabled assessment that supports meaningful, rather than simulated, self-directed learning.

Username
Password