Language teachers’ emotional regulation in generative AI integration: a sequential explanatory mixed-methods study

Abstract

Generative artificial intelligence (GenAI) is among the most consequential recent innovations in language education, yet little is known about how teachers sustain engagement with it over time. Research has documented teachers' initial perceptions of and attitudes toward AI tools, but how teachers continuously regulate the emotional tensions arising as GenAI becomes embedded in everyday professional practice remains underexplored. This mixed-methods study examines emotional regulation as an ongoing professional process through which language teachers negotiate GenAI integration. Using a sequential explanatory design, survey data from 34 language teachers working in varied international settings, from schools to universities and private language institutes, mapped how teachers engaged with GenAI and the emotional tensions this generated. These findings then informed in-depth interviews with seven participants at two time points approximately six months apart, tracing how emotional tensions shifted as professional engagement deepened. Guided by the Language Teacher Emotion Model (LTEM), the findings reveal sustained emotional ambivalence as the defining feature of teachers' experience: the creative possibilities GenAI opened up were persistently shadowed by anxiety, ethical unease, and professional uncertainty. These tensions were seldom resolved through acceptance or rejection; instead, teachers drew on regulatory practices that allowed them to incorporate GenAI selectively while protecting their sense of professional competence and identity. Emotional regulation, in this sense, functioned as a form of active professional positioning. The study reframes GenAI integration as a process of affectively mediated professional adaptation and points to the need for institutional support that attends to the emotional dimensions of innovation in language teaching.