IMPROVING THE METHODOLOGY OF TEACHING ARABIC AS A FOREIGN LANGUAGE TO UNIVERSITY STUDENTS THROUGH GENERATIVE ARTIFICIAL INTELLIGENCE TECHNOLOGIES
The article examines the methodologically grounded integration of generative artificial intelligence technologies into the teaching of Arabic as a foreign language in higher education. The author proposes a model of teacher-mediated three-stage AI mediation, in which interaction with a language model comprises a teacher-designed prompt, the student’s independent dialogue with the model, and a subsequent verification and reflection stage. The model was piloted during the first semester of the 2025/2026 academic year with 55 second-year students. The results demonstrate significantly higher gains in lexical range, oral fluency and learner autonomy in the experimental group. The difference in grammatical accuracy proved less pronounced, which is explained by the language models’ errors in vocalisation and in assigning i‘rab. Methodological recommendations are offered on this basis.