AI Pedagogy in Higher Education: A Scoping Review of Competencies, Enactments, and Ethical Governance

Abstract

Artificial intelligence (AI) is increasingly embedded in higher education, yet the pedagogical field intended to organise its use remains conceptually underdeveloped. This scoping review mapped and synthesised peer-reviewed literature on AI pedagogy in higher education, using teacher education as the primary analytic lens. Following Joanna Briggs Institute guidance and PRISMA-ScR reporting principles, searches were conducted across ten databases and publisher platforms for literature published between 2022 and March 2026. Of 956 records screened, 191 studies were included. Data charting and synthesis combined deductive coding, an a priori codebook, AI-assisted coding in Atlas. ti, and human inductive open coding. Six interrelated themes emerged: AI pedagogy is contextually produced rather than universally defined; conceptualised through pedagogical integration, competence, and borrowed theories rather than a settled standalone theory; enacted mainly through augmentation-oriented practices and generative AI tools; and associated with changing teacher professionalism, ethical mediation, governance, structural constraints, and a rapidly expanding but methodologically uneven evidence base. The review rejects purely technical, universal, or tool-centred models of innovation. Instead, it positions AI pedagogy as a context-specific, human-led, competence-rich, and governance-aware form of teaching in which AI supports planning, feedback, adaptation, and reflection without displacing human educational responsibility, especially within teacher education contexts and programmes.