Most young people aged 13-16 already use generative AI (GAI) tools for various purposes, including homework support and when writing stories, as demonstrated by a large UK survey (Picton & Clark, 2024). Notably, the emergence of GAI poses new challenges for Danish L1 teachers: The new technology has not made the complex act of writing any easier. In fact, more than ever, schools must support students navigating new textual demands and developing a reflexive stance when engaging in hybrid practices, including an awareness of their own voice (Kalantzis & Cope, 2025; Stornaiuolo et al., 2024; Wang, 2024). To this end, there is a significant knowledge gap concerning how people in general – and lower secondary students in particular – use the new tools (Luther et al., 2024). Developing empirical knowledge of hybrid text practices in schools is therefore not only of great scientific relevance, but also considerable societal importance, helping schools and teachers support students’ development as writers and academic well-being through engaged and knowledge-based instruction – today and in the future (UVM, 2025; Simovska & Primdahl, 2024). We anticipate that the use of GAI might affect students’ sense of ownership and involvement in the learning process in ways that we need to better understand if we are to support their writing development, identity formation and participation in and beyond educational contexts. Furthermore, to what extent will students create texts that blend their own and artificial voices? And how might these multiple voices affect textual coherence? To investigate these potential new practices involving student texts, we draw on established key concepts of polyphony (Bakhtin, 1981, 1986) and semantic coherence (Hasan, 1984, 1989), at the same time expecting to enrich these concepts with new dimensions. Thus, the aim of this project is to generate knowledge about human–AI hybrid text practices in schools by examining students’ observable writing processes in terms of strategies and interactions, their texts, and their explicit reflections on these processes and their writerly choices. We focus on Year 8 Danish L1 classrooms, an under-researched stage of school that is highly relevant due to the presence of GAI tools in students' everyday lives and because students are in the process of developing sophisticated literacy competencies (Moje, 2015; Shanahan & Shanahan, 2012). We explore the following research questions:
RQ1 What characterises Year 8 students’ observable writing strategies and interactions when participating in hybrid text practices with generative AI tools?
RQ2 What forms of polyphony and coherence can be identified in texts created by students through hybrid text practices with generative AI tools?
RQ3 How do students reflect on the writing process and their writerly choices when creating texts?
