A Case Study of Chinese University Students Negative Emotions in GAIAssisted L2 Writing

Previous research, predominantly quantitative, has yielded contradictory findings regarding the impact of generative artificial intelligence (GAI) tools on negative emotions experienced by learners in second language (L2) writing. However, little qualitative research has been conducted to further explore these emotions in GAI-assisted contexts compared with non-GAI-assisted ones. This qualitative case study investigated three Chinese undergraduates negative emotions with the assistance of GAI tools (e.g., ChatGPT, DeepSeek) in L2 writing. Using thematic analysis of in-depth semi-structured interviews, the study found a complex, nuanced picture. Though GAI assistance largely alleviated participants anxiety, uncertainty, and frustration, these negative emotions persisted, albeit for different underlying reasons. For example, anxiety about limited L2 proficiency was eased by GAI-generated feedback, yet participants became anxious about the reliability of such feedback. Moreover, new negative emotions, including distrust, anger, and shame, emerged due to the involvement of GAI tools. Pedagogical implications are discussed in light of these findings.