Abstract
This study aims to capture the stance of YouTube video content creators regarding the use of generative AI for language learning. Opinion mining and sentiment analysis techniques were employed to analyse the content, comments, and transcriptions of 66 YouTube videos published from December 2022 to October 2023. The findings revealed that most videos focused on speaking (n = 40) and writing skills (n = 24), with fewer videos addressing listening (n = 3) and reading (n = 19) skills. Sentiment analysis showed that videos predominantly conveyed optimistic (n = 42) and analytical (n = 17) sentiments, indicating a generally positive stance towards generative AI for language learning. Clustering analysis identified four thematic clusters: "language development and practices" (n = 33), "basic expression skills" (n = 25), "intercultural communication skills" (n = 6), and "language structure and meaning" (n = 2), representing different approaches to language learning with generative AI. Cross-sectional analyses revealed fluctuations in video counts and sentiment scores over time, with higher expectations for generative AI in writing and speaking skills, and relatively lower interest in listening skills. The findings suggest that YouTube video creators generally view generative AI as a promising tool for language learning, with a focus on developing practical communication skills, fostering intercultural understanding, and facilitating language development. These insights can inform the design and implementation of AI-supported language learning materials and practices.