Abstract: This paper investigates the effects of facilitation approaches of teachers on the engagement of the students in the asynchronous learning environment mediated by the AI chatbots. Though the chatbots provide the benefits of immediate feedback, personality-based feedback, and constant interaction, the outcome of the educational technology is largely dependent on the facilitation approach of the teachers. With the mixed-methods correlational approach, the study collected the data related to the usage logs of the chatbots, sentiment, scores of the quiz, and facilitation inputs with 100 undergraduate IT majors and 20 teachers. The results show that the engagement of the chatbot (r=.74,β=.45), emotional sentiment (r=.66,β=.33), and facilitation inputs of the teachers (r=.61,β=.29) are all reliable predictors of academic performance, together explaining the collective variance of approximately 65% for the scores of the quiz. The results of the study provide evidence on the complementary effects of human facilitation to optimize the effects of AI-facilitated learning. Furthermore, this study also promotes Sustainable Development Goal 4 (SDG4) to provide inclusive, effective, and quality learning for the users through the collaboration of human and AI resources.