The use of Generative AI (GenAI) tools in writing is a contested area in both the literature and the higher education (HE) practitioner community (eg., “JISC AIED email list”). While studies highlight the widespread adoption of GenAI tools in writing among university students (Alharbi, 2023) and its potential to enhance the quality of written work (eg., Bal & Arseven, 2025), a key question remains insufficiently explored: what is the role and agency of the writer in the writing process, and whether they develop core academic skills such as critical thinking, synthesis, and evaluation.

Current research on GenAI policy and writing has yet to offer a detailed, process-level understanding of what exactly goes on when producing a piece of academic writing and what skills might be gained or lost when GenAI tools are incorporated in this process. Our research aims to address this gap by developing an empirically grounded, process-oriented model of academic writing to inform students’ meaningful and ethical engagement with GenAI tools.

Guided by the “process-oriented model of writing”, developed by Emig (1971) in the 1970s and her contemporaries and successors (eg., Heyes and Flower, 1986; Rommer-Nossek, 2017; Bhat et al, 2023), our research adopts a qualitative, two-stage design: (1) In-depth interviews with 10 faculty members who are both practitioners and mentors of academic writing, and (2) Focus group interviews with a sample of 20 Masters students. Their reflections on their use of GenAI tools in writing inform the further refinement of the model. The study aims to contribute an empirically rich account of academic writing as a process and offers more nuanced guidance on integrating GenAI tools in ways that support their skills.

The presentation at the conference will report on the research methodology and emerging findings, with implications for pedagogy and policy in higher education.