A Scoping Review of AI for Sustainability and Sustainable AI in Higher Education

Synthesis: Daniel, Podgorodnichenko, and Carr (2026) conduct a scoping review of two interconnected yet distinct pathways at the intersection of AI and sustainability in higher education: AI for Sustainability (using AI to achieve sustainability outcomes) and sustainable AI (reducing the direct environmental and ethical impacts of AI itself). As both users and developers of AI, universities are well-positioned to advance environmental and social goals while ensuring responsible use. The review maps how these concepts are defined, the field's evolution, and universities' actual engagement with AI for sustainability — finding promising examples (energy management, climate monitoring, green campus programs) but limited scale and a lack of clear ethical or environmental guidelines, with much research remaining conceptual or small-scale.

Key Findings

1. Two distinct pathways. AI for sustainability (AI as a tool for sustainability outcomes) and sustainable AI (reducing AI's own environmental/ethical footprint) are related but distinct research areas that are often conflated.
2. Promising but limited applications. Universities use AI for energy management, climate monitoring, and green campus programs, but these efforts are limited in scale and often lack clear ethical or environmental guidelines.
3. Conceptual and small-scale evidence. Much of the current research is conceptual or based on small-scale pilots rather than robust, large-scale implementations.
4. Universities are dual actors. As both users and developers of AI, universities must balance advancing sustainability goals with ensuring AI itself is used responsibly and sustainably.

Relevance

This scoping review is the defining survey for the knowledge base's Sustainability concept page — it establishes the two-part taxonomy (AI for sustainability vs. sustainable AI) that organizes the whole concept. It connects to Ethics, AI Governance, and Meta-Analysis and Systematic Review as a systematic map of the evidence base, and to the environmental-impact dimension of sustainable AI.

What this means for practice

• Administrators. Keep the two agendas separate in institutional strategy — AI for sustainability (AI as a tool for environmental, social, and economic outcomes) and sustainable AI (reducing AI's own environmental and ethical footprint) — because conflating them obscures who is accountable for what.
• Institutions. Move green-campus, energy-management, and climate-monitoring applications beyond limited pilots and attach explicit ethical and environmental guidelines to each one.
• Policymakers. Regulate universities as dual actors — both users and developers of AI — and require evidence on both sides of the ledger rather than treating adoption as inherently sustainable.
• Administrators. Commission evaluations of implemented initiatives instead of assuming benefit: most research in this field remains conceptual or based on small-scale pilots.
• Administrators. Fill the evidence gap for the Global South, where almost no research shows how institutions employ and adapt AI for sustainability, rather than importing findings from North American and European contexts.

Limitations

• A scoping review maps the literature without appraising study quality, so this synthesis cannot rank or validate the evidence it reports — the authors state the review is intentionally broad and identifies gaps rather than adjudicating findings.
• Coverage is geographically skewed: most research on AI for sustainability in higher education comes from the Global North, mainly North America and Europe, with little from the Global South apart from a few South African studies.
• Database coverage had to be expanded across Google Scholar, IEEE Xplore, Scopus, ERIC, Web of Science, Environment Complete, and Business Source Complete; the authors acknowledge that no set of databases fully encompasses a field spanning computer science, environmental studies, education, and policy.
• The concepts themselves are emergent and inconsistent across sources, so the definitions mapped here may not stabilize as the field develops.

Citation

Daniel, B. K., Podgorodnichenko, N., & Carr, S. (2026). A Scoping Review of AI for Sustainability and Sustainable AI in Higher Education. Discover Computing, 29(1), 48.