AI-assisted participatory design: a study on the performance of Large Language Models in community renewal

Within the context of urban regeneration, this study investigated the role of Large Language Models (LLMs) in facilitating public participation in community renewal projects. The aim was to assess whether LLMs could improve consensus-making, participant mood, and workshop outcomes compared with traditional face-to-face discussions. The methodology involved two groups of architectural students engaged in structured design tasks. Group R used regular discussions and Group A used LLMs as a tool. The results indicated that Group A demonstrated a higher frequency of consensus-building and compromise, experienced more positive mood changes, and higher satisfaction with the outcomes. These findings suggest that LLMs can enhance participatory design processes, offering a structured platform for idea generation and conflict resolution in community planning. However, limitations such as artificial experimental settings and reliance on students highlight the need for further validation in real-world settings. This research contributes to the growing field of AI applications in urban design by advancing equitable and innovative community renewal practices.