本研究補充了新興技術採用與TMT決策之理論缺口。實務上，建議企業在數位轉 型初期應組建適度跨界的敏捷團隊，避免盲目追求高度多樣化。同時，企業應深化數據共享文化並強化決策收斂機制，以避免高多樣性團隊在動盪環境下因意見分歧而陷入決策癱瘓。;As Generative AI (GenAI) reshapes corporate value creation， its adoption remains a high-uncertainty strategic decision driven by the Top Management Team (TMT). Integrating Upper Echelons Theory (UET) and the Technology-Organization-Environment (TOE) framework， this study examines how TMT diversity—across gender， tenure， function， and education—influences GenAI adoption， while considering the moderating roles of internal "information readiness" and external "environmental volatility."
Based on a sample of 203 Taiwanese listed companies， empirical results reveal that the impact of TMT diversity is non-linear. Specifically， low functional diversity positively drives adoption by facilitating information complementarity， whereas excessive diversity hinders decision-making due to cognitive load. Furthermore， low gender diversity significantly inhibits adoption breadth， likely due to in-group/out-group biases in skewed team structures. Tenure and educational diversity showed no significant effects.
Regarding moderation， "information readiness" serves only as a hygiene factor and does not significantly mitigate the friction within diverse teams. Conversely， "environmental volatility" exerts a negative moderating effect; under high turbulence， teams with high gender and tenure diversity are prone to "faultlines" and the "threat-rigidity effect，" leading to conservative strategic choices and reduced GenAI adoption.
This study addresses theoretical gaps in technology adoption and TMT decision-making. Practically， firms are advised to assemble "moderately cross-functional" teams during early transformation rather than pursuing extreme heterogeneity. Additionally， fostering a collaborative data culture and strengthening decision convergence is essential to prevent high-diversity teams from falling into paralysis under environmental pressure.
