Artificial intelligence in shaping professional vocabulary skills of students in non-language majors

Abstract

Purpose/Aim: This study examines how artificial intelligence (AI) tools can be integrated into English for Specific Purposes (ESP) instruction to support the development of professional lexical competence in students of non-language majors, including psychology, journalism, automotive engineering, and finance. It responds to the need for personalized and context-driven language learning in a digitalized professional environment. Methodology: A qualitative, descriptive study was conducted using a multi-case design. Four ESP courses from two Ukrainian higher education institutions were selected, each representing a different professional domain. Data collection involved classroom observations, analysis of teaching materials, and written feedback from 208 students and 2 instructors. A cross-case analysis was performed to identify AI tools used, including ChatGPT, Gemini, DeepL, Grammarly, Diffit.me, YouGlish, MagicSchool.ai, Anki, Quizlet, Duolingo, Quillbot, and RemNote, AI-based glossaries and simulations, as well as to evaluate students’ responses to them. Results: AI-supported activities were associated with improved vocabulary retention, more accurate use of domain-specific terms, and increased learner motivation. Students demonstrated increased confidence in using professional vocabulary during role-plays, discussions, and written assignments. Educators emphasized the value of clear instructional support and opportunities for reflective application. Reflections revealed both the benefits of AI-driven personalization and concerns about overreliance. Implementation varied by field: simulations were widely used in automotive ESP, while conversational agents and writing assistants were more prominent in psychological and journalistic contexts. Conclusions: AI tools can enhance vocabulary acquisition and learner autonomy when embedded in meaningful, pedagogically guided tasks. The instructor’s role remains essential in selecting appropriate tools, contextualizing content, and promoting critical engagement. The study suggests the importance of ethical integration, domain-specific customization, and further empirical research on AI in ESP teaching.