As artificial intelligence (AI) technology proliferates, its impact on second language education, has garnered significant attention. This study explores the strategies adopted by university students in utilizing AI tools, including machine translation and generative AI, during their second language thesis writing process. By conducting semi-structured interviews with Japanese majors, the study delves into students’ utilization strategies, perceptions, and their desired assistance when applying AI tools for academic writing. The findings reveal that students frequently rely on machine translation for comprehension and initial translation, while generative AI aids in summarizing, outlining, and refining language. While acknowledging the efficiency and quality enhancements AI brings, students also note limitations such as translation quality issues and potential misunderstandings with generative AI. Participants express needs for instructor guidance, training, and recommendations for suitable AI tools. This study concludes that AI tools hold promise for enhancing academic writing but underscores the importance of future research on larger scales to evaluate their long-term effects on writing skills and creativity. Furthermore, developing user-centered AI tools and integrating AI into writing curricula are recommended to maximize the educational benefits of AI-assisted writing.
