This study investigates the integration of Generative Artificial Intelligence (GAI), specifically ChatGPT, into a flipped classroom model for woodworking skill training in engineering education. Traditional woodworking instruction often struggles with limited preparation resources, reduced time for hands-on practice, and the challenge of addressing diverse learning needs. To overcome these issues, ChatGPT was used to generate pre-class learning materials, including conceptual tasks and personalized quizzes, and to provide immediate feedback and automated formative assessments. The research adopted an action research methodology involving 32 university students with varied Kolb learning styles in a semester-long woodworking course. The course was structured in three phases: AI-supported pre-class preparation, in-class skill application, and post-class reflection. ChatGPT-generated content was implemented alongside Google Forms and Moodle quizzes to create an adaptive learning environment. Results from descriptive and inferential analyses show that the AI-enhanced flipped classroom improved students’ pre-class engagement and skill proficiency. Differences in learning outcomes were also observed across learning style groups, suggesting that personalized learning pathways can be effectively supported by generative AI. Qualitative feedback indicated that students appreciated the tailored content and valued the ability to receive on-demand support during their self-study process. This study presents a scalable instructional model leveraging GAI to promote personalized, skill-based education. The findings highlight the potential of ChatGPT to support differentiated instruction and provide insights for designing intelligent learning environments in engineering education.