AI, classification, and education: making up new policy subjects in old and predictable taxonomies

Abstract

This paper explores how artificial intelligence is reshaping classificatory practices in education governance, with a focus on policy subjects such as teachers, students, policy makers, and school leaders. Drawing on Ian Hacking’s concept of ‘making up’ people, we examine whether AI alters the historical role of statistics in constructing policy subjects. We aim to answer two interrelated questions: One, how does AI contribute to the practices of classifying people? Two, what implications does this contribution have for the governance of education? We contend that AI reinforces established logics of classification in education while potentially becoming a new actor in the governing process, influencing policy formation, implementation, and evaluation. Our argument relies on three assumptions: (1) classification is central to the making of policy subjects; (2) AI amplifies existing classificatory mechanisms; and (3) AI may eventually constitute a qualitatively new kind of policy actor in education.