This study attempted Human-AI Co-Creation (HAIC) for fashion image generation to investigate whether AI could provide support and inspiration regarding creative fashion design to human fashion designers, serving as a collaborative partner. As such, the generated fashion images were analyzed and the characteristics of the AI agents were examined, while the research employed both literature review and empirical study methods. The literature review focused on creativity concepts and evaluation tools, HAIC in fashion design, and text-to-image AI, emphasizing relevant literature and previous studies. In the empirical study, an analysis was conducted on the characteristics of the fashion images generated by AI agents and curated by human designers, examining both the resulting fashions and the AI characteristics. The analysis revealed that fashion images generated through HAIC exhibited diverse aesthetics depending on the AI agent, containing unpredictable elements despite using the same prompt. In addition, creativity varied depending on the prompt, but overall elaboration, novelty, and aesthetics were found to be excellent. The options provided by AI agents facilitated fast and fluent modifications and transformations of fashion designs, enabling the curation of human fashion designe...