The use of generative AI in qualitative management research has sparked an expanding methodological debate. This paper conducts a debate review of major scholarly contributions published between 2023 and 2026, organized around five key issues: the epistemological status of LLM outputs, misrecognition mechanisms, interpretive agency and deskilling, efficiency as a scientific value, and the smoothing of participant categories. For each issue, the analysis delineates what is—and what is not—problematic. The findings suggest that the basis for judging the appropriateness of AI use lies not in AI’s technical characteristics alone, but in researchers’ epistemological awareness. Building on this, the paper proposes an evaluative framework comprising three literacies: methodological, technical, and practical. The contribution lies in reframing the binary question “should we use AI?” into the more practically grounded question “under which epistemological commitments, how far to delegate the work to AI, and for what purpose?”
