ChatGPT as a learning scaffold in critical theory education among Bangladeshi students

This paper examines how undergraduate students in Bangladesh have used generative AI, particularly ChatGPT, as an informal learning tool to address the linguistic and conceptual challenges of critical theory. Using interviews, written reflections, and AI interaction logs, the study investigates how students utilise ChatGPT to translate jargon, clarify abstract ideas, and contextualise Western theoretical texts by Marx, Freud, Foucault, and Derrida. Their practices suggest both the teaching potential and epistemic risks of AI-supported learning. While ChatGPT acted as a scaffold that made complex ideas more accessible, it also fostered algorithmic dependence, replacing interpretive struggle with a fluent but uncritical understanding. Framed within debates on critical digital pedagogy and postcolonial education, the study argues that students’ spontaneous use of AI creates a “shadow pedagogy” that fills institutional gaps while transforming the relationships among language, technology, and knowledge. It highlights how AI in the Global South functions as a cultural translator within unequal knowledge environments, offering insights into a more inclusive, critically aware integration of AI in higher education.