Proposing Shampoo Primary Packaging Design Using Image-Generative AI : Visualizing Korean Aesthetic

This study used the image-generative AI tools GPT-image-1.5 and Gemini to produce primary shampoo packaging designs reflecting Korean aesthetics and to compare the outputs under identical prompt conditions. Based on domestic literature, Korean aesthetic attributes were organized into five categories—form, texture, color, motif, and emotion—and then operationalized into an executable prompt framework using ChatGPT 5.2. The same prompt was entered three times for each tool, and the results were qualitatively analyzed. GPT-image-1.5 consistently maintained product-category appropriateness, including the pump dispenser, across repetitions, whereas Gemini showed greater variability, frequently modifying or omitting the pump while emphasizing traditional formal interpretations such as ceramic-like textures. Both tools largely reproduced the intended low-gloss matte finish, low-saturation palette, and traditional motifs/Hangul-inspired elements. Overall, the findings indicate that even with the same prompt, the tools prioritize different constraints—category compliance versus traditional aesthetic emphasis—offering foundational guidance for tool selection and prompt-design strategies in K-beauty packaging practice.