FEATURES OF LARGE LANGUAGE MODEL USE BY NON-PHILOLOGY STUDENTS IN THE PROCESS OF FOREIGN LANGUAGE LEARNING

Abstract

The article presents the results of an empirical study examining the use of large language models by first-year non-philology students during foreign language learning. The aim of the study is to investigate the frequency, patterns, purposes, and motivations behind students' use of large language models who are not majoring in a foreign language. The empirical basis of the study consists of survey results from 51 respondents in March 2026. The collected data were analyzed using both quantitative and qualitative research methods. The survey revealed that ChatGPT and Gemini are the most widely used large language models among students. The paper found that more than one-quarter of respondents use large language models daily for educational purposes. The primary motivation for using these tools is time efficiency and the desire to complete academic tasks more quickly, regardless of subject complexity. Large language models are most frequently employed for vocabulary and text translation, grammar explanation, and the completion of lexical-grammar exercises and tests. However, their potential for developing productive language skills, particularly speaking and listening, remains underutilized. The study also revealed that most students are aware that using large language models may lead to violations of academic integrity principles. At the same time, the findings suggest that students do not always have a clear understanding of the boundaries between acceptable and unacceptable uses of artificial intelligence tools in academic contexts. Based on the results obtained, methodological recommendations are proposed for integrating large language models into foreign language instruction, including the use of Text-to-Speech technologies, elements of the flipped classroom approach, and creative communicative tasks in which large language models serve as supportive learning tools.