Instructor Authority and Generative AI in Japanese Writing Instruction

Abstract

The spread of generative artificial intelligence in higher education has begun to change writing classrooms in ways that are not always immediately visible. In second language education, students now revise and refine texts through AI-supported interaction before teachers ever see the final product. This shift raises questions not only about writing quality, but also about how teachers understand, evaluate, and guide the writing process itself. In Japanese writing instruction, where revision often depends on close attention to context, nuance, and interpersonal positioning, these changes carry particular weight. This study examines how ten instructors of Japanese writing at Hanoi University perceive the growing presence of generative AI in their teaching. Based on survey data and qualitative interpretation, the analysis shows that writing instruction remains deeply dependent on teacher labor, especially in areas that require contextual judgment and individualized feedback. At the same time, many instructors described a growing sense of uncertainty. When revision happens through AI outside the classroom, it becomes harder to determine what a student can actually do alone, and this weakens confidence in conventional forms of assessment. Some participants also felt that the teacher’s traditional role as the central source of correction is becoming less stable. Yet this did not lead them to view AI as a replacement. Instead, they emphasized that meaning-making in Japanese writing still relies on human sensitivity to social context and rhetorical intention. Rather than simply introducing efficiency, generative AI appears to be changing the balance of authority in writing education itself.