With the growing intelligence and automation of vehicles, human-vehicle collaboration has become essential in driving tasks. However, the scenarios faced by driving are often complex and risky. Various risk factors can significantly affect driving behavior and increase the risk of accidents. This study explores how intelligent systems with personalized features can improve the human-machine interaction experience, enhance trust and acceptance in dangerous driving scenarios. The study first conducted desktop research and literature to understand the current personality models and robot personality research directions, and summarized typical dangerous driving scenarios. Then, a co-creation design framework was proposed and applied to the process of design workshops. Based on the strategy of human-intelligence personality matching, a design workshop was conducted in which designers and users collaboratively developed an intelligent cockpit HMI design tailored to dangerous driving scenarios. Finally, the design framework and initial design solutions were evaluated. The results indicate that the proposed co-creation design framework and tools effectively supported the design process, fostered collaboration, and ensured that the final design met user needs. This study presents a novel approach to intelligent cockpit HMI design grounded in specific driving scenarios and provides valuable insights into enhancing user experience and interaction.