Art students' perceptions of the use of AI tools in ESP classrooms

Abstract

This study addresses the following research question: How do students of applied and dramatic arts perceive the use of artificial intelligence (AI) tools in English for Specific Purposes courses, education and artistic practice? Drawing on theoretical discussions of criticial thinking, pedagogical innovation, creativity and ethical AI integration in education, it examines the pedagogical potentials and limitations of AI-assisted learning in higher education. Based on three-phase sequential multi-method design, it combines qualitative and quantitative approaches. Phase 1 included a Multi-Criteria Analysis (MCA) of eight AI tools, used to assess their suitability for ESP instruction. Phase 2 consisted of a quasi-experimental AI-enhanced ESP intervention implemented with four student groups from the Faculty of Applied Arts and the Faculty of Dramatic Arts. Phase 3 questionnaire-guided semi-structured interviews were conducted to investigate students' perceptions of AI use. A convenience sample of 144 firstand second-year students was included. Data were collected using an MCA quality-criteria checklist (κ = .68), corpora of AI-enhanced exercises and prompts, Likert-scale items (α = .858), semi-structured interviews and digital teacher diaries. Quantitative data were analysed using descriptive statistics and t-tests, while qualitative data were examined through a descriptive-interpretive approach. Across the variables of imotivation, creativity, interactivity and language skills development, the findings indicate that the selected AI tools were generally beneficial for structured language practice, but insufficient on their own for integrative ESP learning. Twee, Perplexity and Ginger Software mainly proved useful, whereas TTS tools and Skybox AI by Blockade Labs required additional pedagogical support. All groups repeatedly emphasised the importance of teacher guidance and human interaction in language learning. Although recognising the usefulness of AI for vocabulary, grammar and pronunciation practice, or in writing and repetitive tasks, students were critical regarding reliability, ecological impact, ethical implications and transparency in AI-based work. The future artists largely rejected the use of AI in creative practices, seeing it as a potential threat to artistic identity. These findings underscore the importance of critical thinking, transparency, trust and ethical awareness in AI-enhanced language teaching and education, and contribute to the limited research on AI tools beyond ChatGPT in ESP teaching for artistic purposes.