Neurodiversity and generative AI in higher education: A sequential multi-method qualitative study of students’ interactions, motivations, and strategies

Neurodivergent students have been largely overlooked in research examining student experiences with generative AI (GenAI) in higher education, creating critical knowledge gaps about inclusive educational technology use. This sequential multi-method qualitative study examined how neurodivergent students (NDS) and neurotypical students (NTS) interact with and experience GenAI in academic contexts. Using a sequential design, we employed a virtual revisit think-aloud protocol with 24 university students (12 NDS, 12 NTS) completing an academic task using Microsoft Copilot, followed by focus groups with 14 participants (7 NDS, 7 NTS) to contextualise these task-based behaviours through broader GenAI perceptions. A neurodivergent research advisory board co-designed all procedures. Three themes emerged: approaching the task efficiently, evaluating and refining output, and maintaining academic integrity. While both groups demonstrated strategic GenAI use, neurodivergent participants in this sample explicitly described their cognitive processing needs, using GenAI to manage energy, and humanising some AI interactions. In the neurodivergent focus group, students described GenAI as essential scaffolding for managing academic demands, though this created complex tensions around authenticity and over-reliance. Interface design challenges seemed to particularly disadvantage users with executive function differences. As institutions rapidly integrate these technologies, inclusive design principles and empirical evidence regarding diverse student needs are essential to ensure equitable access to AI-supported education.