This paper proposes a situated approach to AI ethics in education, grounded in Cultural-Historical Activity Theory (CHAT) and ecological systems theory, to conceptualize how ethical decision-making around AI emerges across interconnected individual, institutional, cultural, and geopolitical contexts. Moving beyond universalist and principle-based ethics frameworks, we frame AI as a non-neutral socio-technical assemblage whose ethical implications are historically produced and locally negotiated. The framework is applied to a comparative analysis of 7 national cases (Australia, Finland, England, France, Italy, New Zealand, and South Korea), mapped across five ecological levels: self, classroom/group, state/geopolitical, cultural norms, and global. The analysis reveals recurring tensions between innovation agendas and ethical concerns, including uneven teacher autonomy, regulatory overload, fragmented policy guidance, cultural anxieties about automation, and global pressures related to competitiveness, labour markets, and platform dependence. Across contexts, teachers are frequently positioned as moral gatekeepers of AI use while lacking adequate structural, institutional, and epistemic support to exercise ethical agency. By integrating CHAT’s attention to contradictions, power relations, and transformative agency with an ecological perspective on nested systems, the paper demonstrates that ethical AI practices are not implemented through compliance but negotiated through situated action. We argue that ethical engagement with AI requires context-sensitive, collective, and transformative approaches that extend AI literacy beyond technical skills toward critical, political, and ecological forms of agency.
Raffaghelli et al. (Fri,) studied this question.