There is a newer version of the record available.

Published October 6, 2025 | Version v1

Education is a special sector: Why we need ethical regulation of Artificial Intelligence and Education (AI&ED) and how we benefit from it

Authors/Creators

  • 1. ROR icon University of Bonn

Description

This paper analyses the ethical dimensions for the design and evaluation of Artificial Intelligence (AI) and Education and provides related  recommendations. It starts with a literature review and discussion of the relationship between AI and Education (AI&ED). We present that education is a special sector for three main reasons: Education is (1) a human right, (2) a complex ecosystem and (3) self-learning. Thus, it requires particular rules and regulations for the AI introduction and use, leading to general benefits. Ongoing developments of AI regulations are currently missing a recognition of this important demand and of the vulnerability of the educational sector. Human rights, democracy, and rule of law as the key global values should guide the ethical debate about AI&ED in combination with specific educational requirements. Based on our literature analysis and discussions, we propose to transfer and adapt the three generic learning quality dimensions to the AI&ED design and evaluation leading to the three dimensions: AI&ED objectives, AI&ED implementations and AI&ED results. In future AI&ED research, they can be validated to focus educational purposes for AI selection. Finally, we recommend three ethical imperatives for the AI&ED design and research directions: (1) impact-oriented, (2) technology-independent and (3) society-focused to guarantee positive achievements and benefits by AI&ED for all three educational levels and involved stakeholders. All stakeholders in education should call for broad discussions to gain consensus on AI permits and prohibitions and to facilitate regulation of AI&ED reflecting the requirements of the special sector and our societal future.

Files

Stracke_2025_Ethical_regulation_of_AI_and_ED_final_preprint.pdf

Files (422.5 kB)