This project introduces a novel and rigorous approach to evaluating the human-likeness performance of the AI-generated texts in five Non-Latin Alphabet Languages (Chinese, Hindi, Japanese, Korean and Russian), and its application. We believe that this article not only fills a critical gap in the current literature on AI but also holds significant potential to lay the groundwork for future advancements in AI text generation, particularly in Non-Latin Alphabet Languages. The project team is composed of seven experts in Chinese, Hindi, Japanese, Korean, and Russian languages, and two researchers specializing in GenAI for foreign language teaching. The raw data on the project is in the attached file.