This study introduces CEWill-Auth, an AI-powered dialogue scenario authoring system that addresses critical challenges in language learning resource development. By integrating Large Language Models (LLMs) with structured interaction frameworks, the system enables educators to generate sophisticated dialogue scenarios across different proficiency levels. The research evaluates the system’s potential through interviews with language educators, demonstrating its capacity to create contextually nuanced, pedagogically aligned conversation scenarios. Key innovations include an intuitive authoring interface, interactive dialogue flow visualization, and adaptive content generation that maintains linguistic authenticity while supporting specific educational objectives.