Abstract Research on generative artificial intelligence (GenAI) for second language (L2) writing has raised concerns about voice erosion and language homogenization, despite the promise of human-AI hybrid intelligence. However, while research foci have been on GenAI’s capabilities, the role of authorial identity, defined as a writer’s perception of self as a legitimate author, remains underexplored in high-stakes L2 settings. This qualitative study examines how Chinese high school students negotiate this identity during GenAI-mediated English writing, framing it as a key perspective for understanding human-AI collaboration. Using a five-week multiple-case study with three participants, we gathered multimodal data, including 23 video-recorded writing sessions, GenAI chat logs, drafts, three-round interviews, and think-aloud protocols. Quantitative multimodal analysis shows ongoing changes in students’ engagement patterns, while qualitative narrative analysis uncovers a cyclical process: invested L2 writers’ desires to pursue imagined identities (e.g., becoming a “top student”) shape their interaction with GenAI, forming distinct states of authorial identity that prompt moments of return-investment evaluation (e.g., “Does this text reflect me? Is my effort worth it?”), guiding subsequent engagement with GenAI. Conversely, a learner with low investment bypassed the evaluative process of authorial identity, engaging with GenAI transactionally. Conceptually, we position authorial identity as a negotiated outcome and a vital evaluative lens within the cyclical loop that continuously recalibrates human-AI collaboration. In practical terms, this research provides educators with insights into how to harness students’ imagined future selves and foster authorial identities to support the development of L2 writers in the era of GenAI.
Jiang et al. (Tue,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: