Stimulating or constraining creativity? Traditional vs. generative AI on divergent thinking in product design

The rise of generative AI (GenAI) has brought both opportunities and challenges to product design, raising the question of whether it stimulates or constrains creativity. This study empirically examines the effects of traditional and GenAI-assisted ideation methods on cognitive creativity and perceived creativity support during the divergent thinking stage of product design.

Focusing on both the creativity of design outcomes and designers' self-reported user experience, a mixed-methods experiment was conducted with 41 sophomore students majoring in product design. Three ideation methods were compared: unaided, Internet-assisted, and GenAI-assisted.

The results revealed that, among novice product design students, traditional methods more effectively enhanced divergent thinking performance, whereas GenAI-assisted methods significantly improved designers' perceived creative support, specifically on usability, exploration, and immersion during the ideation process. Moreover, no significant differences were observed among participants with varying levels of creative tendency.

These findings suggest that, although GenAI has not yet surpassed traditional approaches in generating highly original design outcomes, it serves an important facilitative role in design education by enriching learners' creative perception and affective engagement. The study provides empirical evidence for the thoughtful integration of GenAI into the design process and highlights the need for future research on optimizing human-AI collaboration to augment, rather than replace, human creativity in design learning and practice.