The Digital Education Outlook 2026, published by the OECD in January 2026, highlights the risks associated with the educational use of general-purpose generative AI, including a potential decline in learners’ metacognitive engagement, and underscores the importance of developing and utilizing education-specific AI. However, the actual state of educational AI services and the challenges they face remain insufficiently examined. This study attempts to construct a typology of educational AI services by reviewing the OECD framework and relevant prior research. In addition, against the backdrop of rapid advances in vibe coding technologies that are making it increasingly possible for teachers to develop educational applications themselves, the study provides a theoretical examination of the limitations of existing services and the environmental arrangements needed for the future. On this basis, it proposes a framework for conducting vendor interviews.
Research papers (research meetings, symposium materials and others)