The integration of generative artificial intelligence (GAI) tools is transforming workflows, especially in creative domains like design. By enhancing efficiency and providing diverse sources of inspiration, GAI tools have gradually become collaborative partners for designers, fostering co-creative working models [1–3]. However, how design students—the future designers—perceive and adapt to the application of GAI tools in design processes remains underexplored. Using a diary method, this study examined how three design students utilized ChatGPT, a GAI tool, over one month to support their design projects. Through thematic analysis, the study categorized ChatGPT’s contributions to four key tasks: data processing, design thinking, design development, and design evaluation. Despite these benefits, limitations such as information overload, limited creative breakthroughs, and inaccuracies in image generation were evident. To address these challenges, students employed strategies including refining prompts, providing specific instructions, and deleting previous records to enhance the accuracy and usability of GAI outputs. This study highlights the dynamic interplay between students and AI in the design workflow, demonstrating both the benefits and challenges of integrating ChatGPT into the design process. Furthermore, the findings underline the necessity for students to not only collaborate effectively with GAI tools but also to supervise and critically evaluate their outputs. Despite the significant potential of GAI tools to enhance efficiency and problem-solving, the irreplaceable role of human designers in driving creativity and ensuring thoughtful oversight remains vital.