Methodology for developing the Russian communicative competence of Uzbek-speaking students using generative artificial intelligence

Abstract

The article substantiates a methodology for developing the Russian communicative competence of Uzbek-speaking students using generative artificial intelligence. The study is relevant due to the contradiction between the high linguodidactic potential of large language models and the lack of methodological solutions addressing Uzbek and Russian typological differences and cross-linguistic interference. The theoretical framework integrates competence-based, communicative and contrastive approaches, the technological pedagogical content knowledge model and formative assessment. Russian and international research on generative artificial intelligence in language teaching is analysed, revealing a shortage of studies on Turkic-speaking learners and criterion-referenced assessment. The methodology is implemented through a four-stage cycle comprising diagnostic, generative interactive, reflective corrective and assessment stages, governed by the principles of teacher expert control, transparency and academic integrity. It provides contrastive feedback that explains errors through comparison with the learners’ native language system. Diagnostics, assessment and evaluation are differentiated, and the use of analytic rubrics and performance-based assessment is justified. Experimental teaching shows statistically significant growth in the interactive and pragmatic components of communicative competence and fewer interference errors in the experimental group. The limitations of the methodology, critical verification of machine-generated text as a transferable skill and the changing professional role of the teacher are discussed. The findings apply to teaching Russian at Uzbek universities and teacher development.