A Phenomenographic Study of Students’ Conceptions of Their Collaboration With AI in Learning

Educational theories of collaboration are grounded in human-human interaction, which is common in educational practices. Research has revealed the significance of collaborating with AI in education due to certain benefits. However, it creates conceptual confusion and ambiguity when applied to AI because AI does not process like humans. Existing studies prioritize the performance of AI as a tool-based and quantitative dominance over the nuanced qualitative experience of learners. This study addresses a critical gap by moving beyond a tool-based perspective toward a more profound understanding of their collaboration with AI using a phenomenographic exploration. Thus, this study employed qualitative methods to explore the different ways students make sense of their collaboration with AI in learning. This method employed the phenomenography approach to understand their collaboration, adopting a second-order perspective to describe reality. Data were gathered via in-depth, semi-structured interviews with ten undergraduates across different disciplines at six universities in the southern region of Vietnam. Ten codes were identified into two principal categories: instrumental agency and relational co-evolution. The analysis identified two overarching conceptions involved in operational and ontological sense-making, which focus on tool efficiency and learner growth. This research contributes a broad perspective on distributed agency in socio-technical systems, offering practical implications for educational practices that prioritize human-centered oversight and ethical navigation.