ARTIFICIAL INTELLIGENCE IN TEACHER EDUCATION: EVIDENCE, RISKS AND CONDITIONS FOR PEDAGOGICALLY RESPONSIBLE INTEGRATION

Abstract

The spread of artificial intelligence (AI) systems, especially generative tools, has shifted the educational debate from technical possibility to the conditions under which these systems improve or impoverish teaching, learning and teachers' work. This article examines how the incorporation of AI into education changes the demands, competences, practices and responsibilities of teacher education. It is a theoretical-analytical essay supported by a critical narrative review of a purposively selected corpus of 33 sources (systematic reviews, empirical studies, position papers, reference works and regulatory documents), assembled through a declared search and verification procedure. The analysis covers foundations and applications of AI in education, documented benefits and the conditions under which they occur, impacts on teaching, assessment and teachers' work, ethical and institutional challenges, and implications for teacher education. It argues, by inference from indirect evidence, since no study in the corpus directly tests the effect of teacher education, that the benefits reported in the literature depend on pedagogical design and teacher mediation, and that teacher education cannot be reduced to operational training in tools. As a contribution, it proposes an exploratory conceptual model with four formative dimensions anchored in institutional conditions, not yet empirically validated, together with testable propositions.