Unsettling AI literacies: ethics, power and the pedagogies of critical research

Abstract

As GenAI systems have moved from novelty to everyday use in research and learning practices, higher education institutions are increasingly adopting them in the realms of
“literacy”, “responsible use”, and “ethical awareness”. Such framings often assume strategic simplifications: imagining ethics as a checklist of principles and downplaying the infrastructural forces that shape what counts as legitimate knowledge. This paper argues that approaching AI as a neutral “tool” obscures the value regimes and power asymmetries embedded in its very architecture. Drawing on critical & education technology studies, it proposes a reorientation from AI literacy to critical AI socio-technical practices. Such practices have informed the curricular design, pedagogical implementation, and teaching reflections of two core modules within an online MA distance-learning programme. These are pedagogical practices that aim to foreground situated accountability, epistemic plurality, and alternative imaginaries. By centring observations from curriculum design & learning engagements, this paper advocates for an approach to AI in educational research that foregrounds critical AI literacies: pedagogies attentive to power asymmetries, and that inspire the emergent ways learners may reimagine (and sometimes resist) AI’s role in academic inquiry. This allows for reframing ethics as a politics of value production and infrastructural power and informs how educational research can move beyond the friend-foe binary and toward more contestable engagements with GenAI in knowledge-making.