This paper reviews our empirical investigations into generative AI (GenAI) in foreign language (L2) learning and teaching. Validation of large language model (LLM) performance showed GPT’s strong correlation with human L2 writing accuracy assessments, outperforming Grammarly and demonstrating effectiveness in automated essay scoring. However, LLMs achieved only moderate agreement in classifying open-ended responses, underscoring the need for human oversight. I also identified distinct linguistic features enabling accurate detection of AI-generated texts. Classroom studies revealed Japanese EFL (English as a Foreign Language) learners’ preference for ChatGPT in editing/proofreading, and that their “Ought-to L2 Self” significantly predicted AI tool use. Building on these findings, I proposed the Metacognitive Resource Use (MRU) framework for strategic use of language resources. This framework underlies LexiTracker, a web application providing vocabulary and writing feedback to promote metacognitive engagement. This review offers practical and theoretical insights for the effective integration of GenAI in language education.

This paper examines the ethical, legal, and social issues (ELSI) arising from the integration of generative artificial intelligence (AI) in educational settings in Japan. Following the emergence of ChatGPT in late 2022, Japanese educational institutions initiated pilot programs to explore AI applications in classrooms. Unlike traditional AI systems, where students were merely analyzed subjects, generative AI transforms students into active users, thereby expanding their potential risks and benefits. The study analyzes the AI development process, from training data collection to output generation, identifying key ELSI concerns, including bias, hallucination, privacy violations, and copyright infringement. Educational applications raise fundamental questions regarding the nature of teaching, learning, and schools. This study proposes multilevel response strategies: AI company engagement through procurement guidelines and pledge systems, institutional policy development, and enhanced AI literacy training for both educators and students. Responsible implementation requires balancing innovation with ethical considerations to maintain the trusting relationships that are essential to educational environments.

Generative AI (GenAI) is reshaping how learners and educators engage in content generation, task completion, and feedback. While prompting remains central today, this paper anticipates a shift toward a post-prompting era in which AI agents become more proactive, contextual, and multimodal. In this landscape, the challenge lies not in crafting prompts but in shaping learning trajectories with AI as a co-agent. We propose two key constructs: Meta-Task Awareness (MTA), the ability to discern deeper goals, structures, and strategies behind tasks; and the Chain of Learning (CoL), which reconceptualizes learning as an iterative, generative-reflective trajectory rather than a sequence of isolated tasks. For educators, we extend this to the Chain of Learning Design and Evaluation (CoLDE), which highlights how GenAI can support the design of interconnected learning tasks and the evaluation of student outputs to inform future design. CoL focuses on learner agency in navigating with AI support, whereas CoLDE focuses on teacher agency in orchestrating and improving learning through AI-assisted design and feedback loops. These constructs offer a foundation for human-AI co-agency in learning and teaching. Preparing for this shift demands technical fluency, conceptual clarity, pedagogical imagination, and systemic design foresight.

This study examines the use of digital English textbooks in Japanese junior high schools and their impact on improving English proficiency test scores. Digital textbooks enriched with audio and video enable the recording of operation logs during classes and home studies. This study identified learning patterns and their effects on English proficiency test performance by analyzing these logs using learning analytics. Students who spent more time watching audio content during class and home study showed improvements in their English proficiency test scores. This study also emphasizes the importance of home learning, suggesting that tasks such as audio reading recordings can encourage home study and contribute to better test scores.

Many Japanese junior high school students experience difficulty comprehending geometric proofs, particularly their logical structures. To address this issue, we developed a web-based application to support the visualization and construction of proof structures. This study investigated how different learning formats, namely individual and small-group learning, affect both learning performance and user experience when using the application. A classroom-based experiment involving 33 students showed a significant improvement in proof performance in both learning formats, with no statistically significant differences in learning gains. However, students in the small-group condition reported significantly higher levels of usability and empathy. These results suggest that, while the application facilitates learning across different instructional contexts, small-group learning provides an enhanced user experience through peer interaction. The findings highlight the importance of aligning educational technologies with pedagogical strategies and offer practical insights into the effective integration of digital tools in mathematics instruction.

Honest discussions at the junior high school level are difficult, especially in small schools where interpersonal relationships are fixed and tightly knit. To address this issue, this study explored the potential use of virtual student avatars. The avatars provided rebuttals during discussions with a small class of three junior high school students. The results indicated that the students repeatedly refuted avatar opinions, leading to a noticeable expansion and deepening of their thinking. However, when the students were asked about their impressions of the avatars after the class, they reported negative impressions of the avatars, such as “Don’t make me feel like I am right.” Thus, there is a risk of being disliked when one person argues against the opinions of another person who has real feelings.

This study’s purpose was to clarify teacher’s eye-tracking patterns and practical thinking styles during classroom instruction. To this end, one experienced teacher and one trainee, used a wearable camera to record video of their gaze during class. Subsequently, an analysis of the recorded video was conducted with the participants. The results showed that the trainee looked indeterminately at all classroom children and the speaking children numerous times. The trainee exhibited a trend of passive perception of the events occurring in the classroom. Conversely, the experienced teacher directed their gaze toward groups or individual children, with specific intentions, and exhibited significant proactive perception. This study clarified the following points regarding the differential distribution of teachers’ gaze. Specifically, the trainee focused their attention on issues such as the degree of progress in the classroom lesson. Meanwhile, the experienced teacher focused on developing a deeper understanding of their students at lower achievement levels and learning abilities. Moreover, the experienced teacher was aware of issues requiring their attention and attempted to address them accordingly.

This study aims to clarify the relationship of influence between changes in classroom practice and beliefs about teaching and learning among novice university faculty in PFF (Preparing Future Faculty) program. To this end, we conducted a longitudinal study of six participants in a PFF program that introduced active learning. The results revealed cases where both classroom practice and beliefs changed before and after participating in the PFF program, cases where neither changed, and cases where classroom practices changed but beliefs did not. It became clear that there are two types of changes in classroom practice: “autonomous changes” where with changes in their beliefs and knowledge, teachers changed their classroom practices of their own volition, and “heteronomous changes,” where due to external influences such as training and school policies, teachers changed their classroom practice without any change in their beliefs. Further, the results suggest that even for heteronomous changes to classroom practice, where teachers were able to sufficiently recognize the educational effects through support from experts, etc., it became an opportunity to change their beliefs, leading in turn to autonomous changes in classroom practices.