Smart English for Smart Professions: Integrating AI-Powered Tools in ESP Classrooms

Abstract

The rapid evolution of Artificial Intelligence (AI) has transformed language education, especially within English for Specific Purposes (ESP). While AI tools such as ChatGPT, DeepL, and Grammarly are increasingly adopted by students in health sciences, engineering, and economics, structured pedagogical integration remains limited. This study investigates how these tools influence ESP learning across professional faculties at Universitas Adiwangsa Jambi. Using a qualitative case study design, data were collected from 14 sixth-semester English Education students representing three faculties. Data sources included semi-structured interviews and document analysis of students’ translation and writing tasks. Thematic analysis, assisted by NVivo 14, was used to identify cross-cutting patterns in AI tool usage, challenges, and instructional strategies. Findings revealed three primary difficulties: literal terminology transfer, register control, and discourse cohesion. Students predominantly employed a tool-first, revise-later workflow, often without explicit pedagogical scaffolding. Effective strategies such as guided post-editing, back-translation loops, and glossary-building were identified. Participants also emphasized the need for a structured framework encompassing skill–tool mapping, scaffolded reflection, AI ethics, and teacher professional development. Although AI tools offer significant potential for enhancing ESP instruction, uncritical use can lead to miscommunication and reduced genre awareness. The study highlights the importance of integrating AI within reflective, scaffolded learning environments tailored to each professional domain. This research proposes a practical, student-informed framework to guide ethical and pedagogically meaningful AI integration in ESP instruction, addressing gaps in current practices across disciplines.