GENERATIVE ARTIFICIAL INTELLIGENCE AS A TOOL FOR PERSONALISING LANGUAGE LEARNING: ADAPTING MATERIALS TO LEARNERS' PROFICIENCY LEVELS AND INDIVIDUAL NEEDS

This article examines the use of generative artificial intelligence (GenAI) for adapting language learning materials to learners’ proficiency levels and individual needs. Although personalisation has long been regarded as a desirable principle of instruction, its implementation has been limited by teachers’ time and resources. The exploratory pilot study was conducted over six months with 92 students of the Faculty of Power Engineering and Electromechanics, and three cases were analysed in detail: level-based adaptation of technical texts, exercises targeting first-language interference, and formative feedback on writing. The findings suggest that adaptation is most reliable when prompts contain specific, operationalised parameters, whereas a general CEFR label may be interpreted only approximately, occasionally yielding lexis above the target level. Personalisation appears most productive when it responds to learners’ actual difficulties rather than merely simplifying text. The study underlines the changing role of the teacher as designer of prompts and critical evaluator of machine-generated input. The conclusions are exploratory and define hypotheses for a controlled study. Keywords: generative artificial intelligence; personalised learning; language teaching; CEFR; first-language interference; teacher mediation.