Hybrid reading practices: how GenAI summarisers can support rather than replace reading

Generative Artificial Intelligence (GenAI) summarisers promise to assist students by efficiently condensing complex reading materials. However, despite early evidence of substantial adoption by students, there remains a lack of scholarly research exploring students’ actual practices. Accordingly, this study investigates how undergraduate students employ GenAI when engaging with academic texts. Our analysis of semi-structured interviews (n = 24) with undergraduate students in the reading-intensive context of philosophy suggests that rather than universally substituting AI summaries for reading, students are frequently developing hybrid reading practices that integrate AI tools into traditional reading. We document four such practices (orientation, clarifying, translating, cross-checking), showing that while students can and do use AI to avoid reading, students also use these hybrid reading practices to overcome barriers to access that would otherwise prevent engagement entirely. This distinction between ‘shortcuts’ and ‘access’ challenges assumptions about AI's role in education. As institutions develop AI policies based on concerns about academic shortcuts, our findings suggest these policies may misunderstand how students actually use these tools. Students are creating new forms of academic literacy that deserve serious scholarly attention.