Governing Generative Artificial Intelligence in Schools through Artificial Agency, Competence Frameworks and Teacher Professionalism

Abstract

The integration of Generative Artificial Intelligence (GenAI) in schooling cannot be reduced to a merely instrumental update of teaching practices. Rather, it inaugurates a transformation of the epistemic ecology of schooling, reshaping how knowledge is produced, circulated and — crucially — validated, with anthropological, ethical, and pedagogical repercussions. This paper frames AI not as “intelligence” in a human sense, but as artificial agency: an operational power that can act and respond without intentionality, comprehension, or moral responsibility. Such a stance enables a critical interrogation of educational anthropomorphism, automation of judgement, and the hetero-direction of formative processes. Drawing on international principles (UNESCO; OECD) and European competence frameworks (DigComp, DigCompEdu, LifeComp), in dialogue with Italian scholarship on media education and teacher professionalism, the article argues that the pivotal issue is not “adoption” per se, but the didactic translation of ethical and competence architectures into situated practices: assessment, design, inclusion, relational care, and data governance. The central thesis follows: for technological innovation to remain humanly intelligible and educationally justifiable, teacher professionalism must be strengthened as a projective, evaluative, and deontological function — capable of embodying an ethics that is not merely declared, but enacted through everyday decisions and institutional routines.