The emergence of academic writers: Multilingual doctoral students’ translanguaging and transpositioning in AI-mediated academic writing

This study explores how multilingual PhD students navigate AI-mediated academic writing and construct their academic identities, drawing on translanguaging and transpositioning as analytical frameworks. While generative AI tools are increasingly integrated into academic writing, current research often overlooks how interactions with these tools shape students' roles as emerging academic writers, particularly doctoral students who are in the process of developing their disciplinary voice. Using a cross-context case study, this study analyses data from ten PhD students from diverse disciplinary and national backgrounds. Data include student texts, ChatGPT interaction histories, and interviews, analysed using Interpretative Phenomenological Analysis. Findings show that students strategically mobilize their full linguistic repertoires alongside AI-generated text, engaging in bilingual mediation, multimodal composition, and critical adaptation of AI language. We argue that AI-mediated academic writing is not merely a technical or text production process but an interactive space where these students with insecure ESL stances engage in transpositioning, encouraging the emergence of critical academic writer identities . This process involves students’ power negotiation with AI through translanguaging practices to assert text ownership and disciplinary voice. This study contributes to AI and academic English education research by highlighting identity construction as central to writing with AI and offers pedagogical insights into how doctoral training can better support reflective, genre-aware, and multilingual academic writers in the AI era.