Artificial Intelligence, Power, and Ethics in Sustainability Education: A Systematic Review of Moral Risks, Governance Failures, and Leadership Mediation

Purpose This study critically examines the ethical implications of artificial intelligence (AI) in embedding sustainability competencies within higher education. Moving beyond techno-instrumental perspectives, the study reconceptualizes AI as a normative infrastructure that reshapes moral responsibility, epistemic authority, and governance in sustainability-oriented education. Design/methodology/approach A systematic literature review was conducted following PRISMA 2020 guidelines and guided by the PICo framework. An ethically informed screening process resulted in the inclusion of 75 peer-reviewed studies. Data were analyzed using a hybrid ethical–thematic content synthesis integrated with a higher-order SWOT (HT-SWOT) analysis, foregrounding governance, accountability, and epistemic justice. Intercoder agreement procedures were employed to enhance analytical rigor and analytical transparency. Findings The synthesis identifies three systemic ethical risk pathways associated with AI-enabled sustainability education: (1) algorithmic authority leading to moral deskilling among educators, (2) ethical opacity resulting in accountability diffusion, and (3) epistemic injustice through the exclusion of pluralistic, local, and Indigenous sustainability knowledge. These risks cut across disciplinary and geographical contexts and are mediated less by technological sophistication than by institutional leadership and governance arrangements. Social implications The findings reveal that uncritical adoption of AI in sustainability education risks reinforcing social inequalities, epistemic hierarchies, and exclusionary knowledge regimes, particularly affecting marginalized students, communities, and Global South contexts. Conversely, ethically governed AI has the potential to support more inclusive forms of sustainability learning by amplifying participatory governance, protecting epistemic diversity, and strengthening educators’ moral agency. The study highlights the broader societal responsibility of higher education institutions to ensure that AI-enabled sustainability initiatives contribute to social justice, democratic accountability, and equitable futures rather than reproducing existing structural disadvantages. Practical implications The study underscores the necessity of ethical-by-design governance frameworks in higher education, including transparent accountability structures, leadership capacity building in AI ethics, and institutional safeguards for epistemic pluralism. Without such measures, AI deployment risks undermining the ethical and social foundations of Education for Sustainable Development and SDG4. Originality/value This study makes an original contribution by shifting the analysis of AI in sustainability education from impact-oriented and techno-optimistic narratives to a normative, governance-centered perspective with explicit social justice implications. By empirically grounding AI ethics within sustainability-oriented higher education, the study advances ethical theory, socially responsive policy insights, and institutional practice within the AI and Ethics discourse.