Interrogating AI in education: when peeking inside the black box is not enough

Artificial intelligence (AI) is becoming a pervasive aspect of higher education. It is embedded in digital infrastructures that seek to automate pedagogical and administrative functions, and lately it has extended its reach into the sphere of subjective agency, with generative AI offering countless efficiencies and automated affordances directly to end users. The pervasiveness of AI raises ethical and political questions which concern its dependency on surveillance, the monopolistic tendencies of the companies that develop and govern models and infrastructures, and the social and cultural harms of biased algorithms. This presentation discusses ways in which AI in education can be examined as part of a critical empirical project. It will engage with the notion of ‘algorithmic interrogation’ as a methodological orientation that can assist a more informed lay person understanding of AI, and thus support ethical decision making at different levels, from the institutional level to the individual one.