GENERATIVE NEURAL NETWORKS IN HIGHER EDUCATION: DIDACTIC POTENTIAL AND ETHICAL DILEMMAS

Abstract

Introduction. This paper examines the uses, potential, and limitations of artificial intelligence (AI) in education, with a focus on foreign language teaching. Methods. The study analyzes current practices and standards for using AI in universities in Russia and Kazakhstan, reviews specialized AI tools, and examines data systems from empirical research in digital didactics. Results. The key benefits of AI were identified, including the automation of routine tasks, personalized learning, and the generation of educational materials. Key risks were analyzed, including threats to academic integrity, reduced student mental workload, the development of an «illusion of competence», data unreliability, and legal and ethical challenges. An overview of AI tools used to develop language competencies was presented. Ways to integrate AI through innovative assessment forms and improving the digital literacy of teachers are proposed. Particular attention is paid to the Speechify and ElevenLabs tools for developing language skills. Conclusion. The authors conclude that a balanced approach is needed, combining pedagogical innovation with the preservation of students' intellectual independence. The authors propose ways to integrate technology into education, particularly through the development of unique assignments resistant to automated processing, improving teachers' digital literacy, and the use of AI in mandatory assessment of student work.