Sustainable AI-Supported Reflective Learning in Higher Education: An Integrative Review and the SAREL Design Framework
Year: 2026
Type: article
Abstract: Educational technologies are often described as sustainable because they extend access or improve efficiency, yet sustainability also depends on enduring learning, inclusion, rights, institutional feasibility, and resource use. This concept-driven integrative review examines how generative artificial intelligence (GenAI) can support reflective learning in higher education while preserving learner reasoning and agency. We interpret 49 selected scholarly and policy sources across five conceptual domains: AI-enabled educational practice, reflection and self-regulation, inclusive design, ethics and governance, and environmental and institutional sustainability. Rather than estimating the prevalence of study outcomes, the review identifies the need to consider these interdependent dimensions together. We propose SAREL (Sustainable AI-Supported Reflective Education and Learning), an iterative five-stage framework: specify learning purpose; allocate agency and accessibility; require reflection and verification; evaluate learning, equity, rights, workload, and resources; and loop evidence into redesign. SAREL’s contribution is the integration of established concerns into a task-level design-and-audit logic, not the discovery of new standalone principles. Its rubric and sustainability-adjusted learning value construct are provisional, have not been empirically validated, and require reliability and effectiveness testing. Sustainable AI use is therefore posed as a comparative design question, not an assumed benefit of adoption.
Source: Sustainability
DOI: 10.3390/su182211812
Language: English
Subfield: Education
Field: Social Sciences
Domain: Social Sciences
SDG: Quality education
Open Access status: gold
Estimated APC paid: $2,896
OpenAlex ID: W7237249804
