Responsible AI in Education
Before introducing AI, ask whose learning it should support and according to which values.
Generative AI and educational technology may transform lesson preparation, learning support, assessment and administration. They also raise new questions about educational purpose, teacher expertise, learner autonomy, data, equity and accountability.
The Kano Lab asks not only whether a tool works or saves time, but whether it should be used, who should decide, and how benefits and risks are distributed.
Teachers and AI｜Keeping professional agency at the centre
Teachers should not become people who simply implement AI outputs. Their ability to interpret curricula, understand learners, exercise judgement and choose not to use AI must be respected.
We study how AI reorganises teachers’ work, whether decisions can be explained, and how technology affects relationships between teachers and learners.
Educational data and governance
Educational data may include learning histories, behaviour, assessment and personal attributes with long-term implications. Institutions should not collect everything that can technically be collected. Purpose, data minimisation, retention, access, third-party sharing, deletion and routes for contesting decisions must be designed.
We examine responsibilities among schools, education authorities, universities, companies, families and learners.
ELSI and RRI
Ethical, legal and social issues and responsible research and innovation should not be checklists added after harm occurs. They should shape development, procurement, introduction, professional learning, evaluation and review.
Relevant issues include privacy, intellectual property, bias, transparency, explainability, responsibility, children’s rights, accessibility and environmental impacts.
Inclusion and equity
AI designed for an “average” learner may disadvantage people across disability, language, culture, gender and socioeconomic circumstances. We investigate possibilities for support technologies while asking how to avoid new forms of exclusion, surveillance and dependency.
Key research questions
- How should roles between teachers and AI be designed around educational purpose rather than efficiency alone?
- How can teachers and learners challenge an AI-supported decision?
- What should schools examine when procuring educational data or AI services?
- How does generative AI affect assessment, authorship, creativity and learner self-efficacy?
- How can AI support deaf and hard-of-hearing or multilingual learners without creating additional risks?
- How can educators, government, companies, researchers and the public co-design responsible introduction?
Selected outputs
Kano, K., & Mizumachi, E. (Eds.). (2026). その取扱いで大丈夫？ 教育データ利活用のリテラシー：生成AI時代のEdTechが問う学校教育のあり方 [Are you handling it responsibly? Literacy for the use of educational data]. Nakanishiya Shuppan. Publisher page
Goto, T., Horiguchi, G., Kanzaki, N., et al. (2024). Balancing human teachers and AI in education: A discussion paper from ethical, legal and social perspectives. OECD Future of Education and Skills 2030/2040. Official PDF
Wakabayashi, K., Sato, H., Takahashi, S., & Kano, K. (2024). 教育データEdTechの導入とELSI対応のグローバル動向に関するインタビュー記録 [Interview record on global trends in educational-data EdTech and ELSI responses]. ELSI NOTE, 47, 1–32. DOI / official page
Kano, K. (2023). ぺた語義：教育データ利活用EdTechへの期待とELSI [Expectations and ELSI for educational-data EdTech]. Information Processing, 64(7), 332–336. DOI / official page
Kano, K., Kanzaki, N., Kishimoto, A., et al. (2024). 教育データ利活用EdTechのELSI論点フレームワーク [An ELSI issue framework for educational-data EdTech]. In プラットフォームと社会基盤：How to engage the monsters. Keio University Press. Book information
Recognition and international contribution
- 2022: Third place, Startup Pitch@CIC EdTech, The Future of Learning with Technology
- Contributions to international discussion related to the OECD Teaching Compass
- International conferences and expert activities connecting education, AI and science communication
Research opportunities
We welcome proposals from education, AI ethics, information studies, law, policy, data governance, teacher education, disability studies and design. Proposals should not assume that a particular technology must be introduced; begin with educational purpose and the people affected.