Generative AI and Pedagogical Sustainability in Higher Education: A Critical Integrative Review with a European Application

Generative artificial intelligence (GenAI) promises personalized support, feedback and efficiency in higher education, but raises concerns about cognitive dependency, unequal capability, assessment validity, privacy, environmental cost and commercial dependence. This critical integrative review draws on a purposive, criteria-based selection of 77 sources from Scopus, Web of Science, ERIC and European Union and European Higher Education Area (EHEA) policy portals, classified by type, context and evidence level; key evidence syntheses were appraised with AMSTAR-2 or the JBI scoping review checklist. Meta-analyses report moderate positive average effects on achievement, motivation and higher-order thinking (g ≈ 0.45–0.80), but heterogeneity is high, very large language-skill effects are probably inflated by small studies, evidence of metacognitive benefit is weak, and durability is untested. We define pedagogical sustainability as the production of durable, transferable learning under economic, social, environmental and institutional conditions that can be maintained over time. We identify five tensions and propose the Sustainable GenAI Pedagogy Framework (SGPF), in which demonstrable learning value is a necessary condition rather than a tradeable pillar. We compare it with eight frameworks and operationalize an Institutional GenAI Maturity Model with indicators, scoring rules and an illustrative application. The framework is general in design and calibrated to European regulation.