ESG 지향 AI 페르소나 구축을 위한 극단적 사용자 참여형 디자인 프레임워크

AI technologies are increasingly used in participatory design to support decision-making and address challenges related to social responsibility. This study proposes a method for generating ESG-oriented virtual extreme users (AI Personas) using AI-based semantic modeling and integrating them into participatory design. The method aims to examine how cognitive–behavioral alignment and misalignment in sustainable user actions influence the gap between organizational ESG strategies and actual user behavior.Based on four semantic dimensions of ESG—Behavior, Data, Method, and Context—the study combines semi-structured interviews, social data analysis, and AI semantic modeling to construct two personas: a Positive Persona with strong ESG alignment and a Negative Persona with notable cognitive–behavioral inconsistency. A textual similarity analysis between AI personas and real users was conducted to verify reliability and reproducibility.Results show that the AI personas capture key variations in user values and behaviors and can be used as analytical tools for exploring sustainability-oriented design scenarios. The study contributes a methodological approach for embedding value-based user requirements into early participatory design processes. Limitations include a small validation sample and constrained data scope, suggesting the need for broader quantitative studies in future work.