Аннотация
The rapid advancement of artificial intelligence (AI), particularly generative AI tools such as ChatGPT, has fundamentally reshaped English Language Teaching (ELT) practices worldwide, yet the intellectual landscape of this research domain remains fragmented at a global scale. This study aims to map the development, intellectual structure, and emerging thematic trends of AI research in ELT from 2018 to 2025. A bibliometric analysis combined with Latent Dirichlet Allocation (LDA) topic modeling was performed on 1,622 articles retrieved from Scopus and Web of Science following the PRISMA 2020 protocol. Performance analysis and science mapping were conducted using VOSviewer 1.6.20 and the Bibliometrix R-package, while LDA topic modeling was implemented in Python (Gensim 4.3). Results reveal an exponential growth of publications, with a 168% increase following the release of ChatGPT in November 2022. The United States, China, and the United Kingdom dominate productivity, while Indonesia and Kazakhstan emerge as growing contributors within the Asian collaboration cluster. LDA identified eight dominant topics (k=8, C_v coherence score=0.61), with "Generative AI and Writing Feedback" and "Chatbots and Conversational Practice" representing the most active post-2022 research fronts. The study contributes a comprehensive intellectual map of AI-ELT research and identifies underexplored areas as priorities for future research