The democratic risks of AI in education: an epistemic perspective

The rapid deployment of Artificial Intelligence (AI) in education poses challenges and risks that are still largely under-researched. To conceive of AI as an ‘epistemic technology’ is useful to look more closely into these risks. The recent epistemic approaches to the study of injustice and democracy, and the tradition of critical AI studies have much to offer in support of this project. Those contributions are useful to outline a critical research agenda for the systematic study of the epistemic dimensions of the risks of AI in education and to make this technology safe for democracy. To achieve this, we suggest combining these approaches with the idea that AI is an institution within an institution, a force that shakes the democratic function of education, and ultimately erodes the robustness of the educational institution.