Critical digital literacies, generative AI, and the negotiation of agency in human-AI interactions

Drawing on the notion of critical digital literacies (CDL) ( Darvin, 2017 ; 2025a ), this qualitative case study investigates the extent to which multilingual secondary school students in Canada exercise agency when they interact with generative artificial intelligence (GenAI) technologies. Recognizing how learners have unequal access to resources that involve the use of GenAI, it poses the question: To what extent do learners enact CDL, and what enables or constrains their enactment? Data were collected through interviews, observations of learners' GenAI usage, and digital artifacts, including screenshots of prompts and AI-generated texts. Thematic analysis was employed to understand learners’ dispositions towards these technologies, and their GenAI practices, while multimodal analysis provided insights into how learners negotiated their agency as they interacted with GenAI platforms. Represented by two exemplar cases, findings show how learners who used GenAI primarily through apps in mobile phones or chatbots embedded in social media platforms, tended to adopt a passive stance towards GenAI, accepting generated output without much critical evaluation. In contrast, learners with access to a laptop and multiple GenAI platforms, demonstrated a more critical disposition towards this technology, comparing outputs across platforms, questioning their reliability, and verifying information through lateral reading. By drawing attention to how material conditions of access and platform design can constitute dispositions towards GenAI and CDL, this study calls on educators to support learners in identifying and challenging inequalities and ideological biases embedded in these technologies, to foster more equitable and inclusive GenAI literacy practices.