The rapid proliferation of Generative Artificial Intelligence (GAI) is profoundly reshaping the educational landscape (Popenici & Kerr, 2017). Existing research confirms AI’s potential in supporting instruction, curriculum design, and assessment through applications such as adaptive learning, personalized tutoring, and intelligent profiling/prediction (Wang et al., 2024). However, while these tools are widely adopted and prioritize efficiency, scholarship overlooks how teachers engage in critical and negotiated interactions with GAI, especially in tasks demanding high artistry, emotional expression, and creative pedagogy like music curriculum design. A systematic understanding of the dynamic equilibrium between professional agency and technological enablement in this specialized context remains a crucial research gap.This ongoing study aims to investigate the dynamic cognitive-behavioral coordination mechanism between master’s students in Music Education at a Macau university and GAI (Deepseek) during curriculum co-design, addressing how this mechanism facilitates the effective balance and adjustment between their professional agency and technological enablement. We employ an Explanatory Sequential Mixed Methods Design (Creswell et al, 2003; Creswell & Clark, 2017) across two phases. Phase One involves purposive sampling of 68 master’s students. Their authentic conversation data with GAI will be meticulously coded using the Toolkit for Systematic Educational Dialogue Analysis (T-SEDA, 2023). Subsequently, Lag Sequential Analysis (LSA) (Bakeman & Gottman, 1997) will be applied to the coded data to identify significant behavioral sequence patterns—specifically those linking teacher instruction behaviors with AI generative responses. This analysis will then be used to categorize core interaction types. Phase Two utilizes retrospective in-depth interviews with selected participants to explore the underlying professional decision-making cognitions, strategic considerations, and contextual factors that inform these patterns.The core theoretical contribution expected is the offering of a unique, behavioral-temporal perspective for understanding how master’s-level music educators achieve this critical balance through negotiated interaction. The preliminary findings will provide valuable empirical insights and theoretical directions for master’s training in music education, curriculum reform, and the future optimization of GAI tools specifically tailored to support music educators.