As the integration of AI into design practice becomes increasingly prevalent, understanding its influence on designers’ thinking is important. However, research on the use and collaboration of AI in design fields is somewhat focused on generative AI itself, such as usability and performance, and there is relatively little research on its impact on users. This research investigated the effect of generative artificial intelligence on designers' thinking during the early design phase of architecture, examining how AI affects designers' divergent thinking and thinking shifts. Divergent thinking is a crucial component of creative thinking, playing a vital role in exploring diverse possibilities and broadening the scope of thought. Three men in their 20s, skilled in hand sketching, CAD, and 3D modeling, performed tasks of similar difficulty in both sessions, both with and without AI. The entire process was recorded and analyzed using a customized evaluation matrix. The results of this research confirmed that generative AI has a significant impact on designers' thinking patterns and divergent thinking. These results suggest that generative AI has the potential to meaningfully support divergent thinking, a key component in the early stages of architectural design. These findings demonstrate that generative AI can have a positive impact on divergent thinking, a skill essential for architectural design, education, and future research utilizing AI.