The purpose of this study is to design a participatory framework for voxel-LED public art that integrates artificial intelligence for sustainable urban engagement. Using a Research through Design (RtD) approach, the framework combines a voxel-LED installation with an interactive kiosk, enabling citizens to co-create content through generative AI and to experience adaptive responses based on environmental data. The proposed system indicates both technical feasibility and design potential, offering two core scenarios: generative co-creation and environment-aware adaptation. While still conceptual in scope, the results specify a replicable model for participatory urban art and provide a roadmap for future studies to evaluate user interaction and advance placemaking-oriented practices.