Learner emotions in AI-assisted English as a second/foreign language learning: a systematic review of empirical studies.

Introduction: The use of generative AI tools in language learning has attracted increasing academic attention. However, previous reviews rarely focus on learner emotions in AI-assisted English as a Second/Foreign Language (ESL/EFL) learning, a crucial aspect for second language (L2) development. The present study addresses this gap by examining 37 relevant empirical studies, including four pieces of grey literature, focusing on the characteristics of recent research on learner emotions in AI-assisted ESL/EFL learning, the emotional variables, and the influence of AI tools on students' emotions. Methods: The researchers conducted a database search in Scopus, Web of Science, OpenGrey, and Google Scholar for relevant articles, following the PRISMA statement. Finally, 37 journal articles were selected in this study. Results: A review of findings indicates that AI tools have been widely utilized in more than 10 non-English-speaking countries, encompassing various English language skills, which suggests that AI-assisted ESL/EFL learning has gained popularity in L2 learning. Both positive and negative emotions were researched, including enjoyment, anxiety, boredom, interest, and shyness, as well as the relevant emotion regulation. Additionally, AI-driven assessment is a rapidly growing research trend. AI-assisted ESL/EFL learning can trigger students' positive emotions and improve their English skills. Discussion: The results of this study provide a deeper understanding of the application of AI tools in ESL/EFL learning and their impact on the emotions of L2 learners in recent years. It will be useful for educators and researchers seeking to understand and evaluate AI-assisted ESL/EFL learning, as well as learners interested in using AI tools.