‘Prior Attainment Affects Uptake of genAI-Based Learning Activities’ with Dr Barbara Piotrowska, Lecturer (Assistant Professor) in Public Policy, Department of Political Economy, King's College London

15 October – 2.30 to 4.30pm

While genAI is often framed either as a shortcut that undermines learning or as a tool for cognitive augmentation, many studies estimating this effect overlook how students with different academic backgrounds engage with genAI when given equal access. Using data from a 2025 post-graduate public policy module that embedded structured, incentivised weekly genAI-based tasks and a genAI-enabled assessment, I examine who engages with genAI and how this engagement relates to academic outcomes. Students were introduced to a variety of uses for genAI over 12 weeks, with relative flexibility in how often they engaged. The findings show that prior attainment strongly predicted engagement: higher-achieving students were more likely to attempt more tasks, raising concerns that even structured genAI interventions may reinforce existing disparities if participation remains uneven. GenAI engagement had no independent effect on final essay or policy brief grades once prior attainment was controlled for. However, it significantly improved performance in a process-focused reflection exercise, suggesting gains in metacognitive awareness.

These findings highlight the value of regular exposure to genAI for developing reflective competence, while also pointing to the limitations of voluntary or lightly incentivised engagement for achieving equitable learning outcomes.