AI-Assisted Error Awareness: Investigating Language Error Recognition and Correction among Japanese EFL Elementary Students

<em>This study examines how Japanese elementary-level EFL students identify and correct language errors through AI-assisted revision. Using a multi-stage approach, students first produced natural language samples describing their hometowns and admired figures, establishing baseline communication patterns. Following the introduction of AI tools, students revised their original submissions and reflected on their learning through surveys. Analysis revealed that students primarily identified and corrected grammatical structure errors, word choice limitations, and sentence complexity issues through AI feedback. The findings suggest that AI-assisted revision enhances students’ metalinguistic awareness and provides opportunities for autonomous learning. This research contributes to understanding how AI tools can be effectively integrated into EFL classrooms to support error correction and language development. Importantly, the results imply that structured integration of AI feedback can foster greater learner autonomy, encourage reflective self-correction, and serve as a scalable complement to traditional teacher-led instruction—especially in contexts like Japan where passive learning tendencies may limit engagement with form-focused feedback.</em>