Freirean Perspectives on AI in Education: Emancipation, Knowledge Creation, and Autonomy

Abstract

With a promise of efficiency-related benefits, Artificial Intelligence (AI) systems are being rapidly integrated into educational practice, be it officially or not. Yet, a growing corpus of research highlights epistemic, pedagogical, and sociotechnical risks, such as bias, hallucination, surveillance, dependency, and the risk of superficial or homogenized learning. Rather than treating these issues as consequences of AI alone, this study examines how AI use reveals longstanding contradictions within result-oriented models of schooling. Drawing on Paulo Freire’s critique of banking education and notions of algorithmic authority, we investigate how AI reshapes autonomy, knowledge construction, and agency in learning environments. We conducted a participatory workshop with students and teachers, analyzing everyday uses of AI and responses to provocative learning scenarios. A theory-informed thematic analysis identified three interrelated contradictions: (1) Control vs. Emancipation; (2) Efficiency vs. Knowledge Creation; and (3) Autonomy vs. Automation. We propose design and subversion guidelines, including Freirean AI literacy and participatory AI design, to support autonomy, critical consciousness, and more dialogic, emancipatory educational practices.