Systematics review on artificial intelligence chatbots and ChatGPT for language learning and research from self-determination theory (SDT): what are the roles of teachers?

Recent advances in artificial intelligence (AI) give chatbots, e.g. ChatGPT, more human-like interaction and conversational capability. AI chatbots are becoming more popular for supporting language learning. Most current review research disregards the significance of teachers' roles in chatbot-assisted language learning. Self-determination theory (SDT) explains the roles by suggesting how teachers satisfy student needs and how chatbots thwart those needs. Therefore, this systemic review aims to: (i) suggest the roles that teachers play in student English learning with AI chatbot; (ii) discuss how those roles satisfy SDT needs of the students; and (iii) discuss the challenges in this learning. This review selected 23 articles published throughout the last ten years (2014-2023). The findings offer (i) four empirical and theoretical contributions: technological pedagogical content knowledge (TPACK) for chatbots, needs satisfaction from teachers, needs thwarting from chatbots, and SDT-based research designs for chatbots; (ii) two practical suggestions: understanding technological knowledge of chatbots and SDT needs support for chatbots; and (iii) six research directions: experimental studies, student language proficiency, more teacher roles, revisiting TPACK, future development of chatbots, and large language model teacher-like chatbots. Overall, the findings enhance our knowledge of TPACK and teacher digital and AI competences for chatbot-assisted language learning.