Integration of Artificial Intelligence into English Language Teaching in Higher Education Institutions: The Potential of Personalized Learning

Abstract

Abstract: Objective. This study is aimed at providing a theoretical substantiation and analysis of the didactic potential of chatbots based on generative language models, in particular ChatGPT, in the process of English language teaching in higher education institutions within the context of personalized learning, taking into account the heterogeneity of students’ language proficiency levels according to the CEFR scale, as well as at identifying the related limitations, risks, and ethical aspects. Methods. The study employs methods of theoretical analysis to generalize Ukrainian and international experience in exploring the possibilities of artificial intelligence in foreign language teaching, as well as a comparative analysis of pedagogical approaches to the integration of chatbots into the instructional process with regard to different levels of English language proficiency. Analytical modelling is used as a research tool to examine the relationship between students’ language proficiency levels, the nature of content personalization, and the expected didactic effect. Results. The findings indicate that chatbots based on generative language models are capable to provide personalized approach to English language teaching through the adaptation of lexical and grammatical material, learning pace, feedback formats, and communicative tasks to individual learner needs. It is demonstrated that the didactic potential of such tools correlates with learners’ language proficiency levels. The study substantiates the feasibility of using chatbots as a means of compensating for the negative effects of mixed-ability academic groups without dividing students into subgroups. At the same time, potential risks associated with the uncontrolled use of artificial intelligence tools are identified, including the narrowing of learning trajectories, fragmentation of language competence, and violations of the principles of academic integrity. Conclusions. It is concluded that the effectiveness of personalized English language teaching supported by chatbots is determined not by the mere use of artificial intelligence, but by the pedagogically sound nature of its integration into the instructional process. Chatbots should be regarded as auxiliary tools for expanding the teacher’s methodological repertoire, provided that the leading role of pedagogical guidance is preserved and ethical standards and principles of academic integrity are strictly observed.