Abstract This article introduces the “dialogic triad,” a model that redefines human-AI collaboration in scholarly writing by integrating relational agency and Bakhtinian dialogism. Through a case study involving two scholars and multiple GenAI systems, we demonstrate how AI serves as an active co-creator, provoking critical reflection and epistemic innovation. The study highlights moments of tension, where AI outputs conflict with human expertise, and synergy, where AI’s generative potential catalyzes theoretical breakthroughs. Crucially, our findings emphasize that human oversight remains central to ensuring accuracy, ethical accountability, and epistemic depth in AI-assisted scholarship. The model advocates for iterative, transparent collaboration that leverages AI’s potential while preserving scholarly rigor. Practical strategies, including iterative prompting, staged engagement and dialogic trace documentation, emphasize the need for human authority in framing arguments and validating AI outputs. Ultimately, this work positions human expertise as the essential anchor for meaningful, ethical, and innovative academic co-creation.
Cao et al. (Wed,) studied this question.