Didactic potential and limitations of generative artificial intelligence in teaching Russian

This article examines the didactic potential and limitations of generative artificial intelligence in teaching Russian as a foreign language. Its aim is to classify instructional scenarios and identify conditions for pedagogically appropriate implementation. The materials include scholarly and methodological publications, UNESCO guidance, and illustrative activities designed to develop grammatical, lexical, and communicative competence. The study employs literature analysis, comparison of teaching approaches, and qualitative evaluation of learning situations. Discussed applications include exercises on case government and verb forms, simulated dialogues, preliminary feedback on written responses, and support for independent practice. Potential limitations include inaccurate explanations, stylistically inappropriate suggestions, insufficient sensitivity to cultural context, and reduced learner autonomy. The article proposes a six stage instructional sequence: independent task completion, artificial intelligence assistance, critical checking, teacher feedback, revision, and renewed independent performance. It also emphasizes personal data protection and expert review of automatically generated recommendations. The research is theoretical and methodological; no classroom experiment or measured learning outcomes are reported. The framework combines digital assistance, teacher guidance, and independent practice. Future classroom validation remains necessary.