Teacher-Mediated AI-Supported Task-Based ESP: Engagement and Perceived Sustainability Awareness among Engineering Students

Teaching English for Specific Purposes (ESP) requires teachers to address learners’ disciplinary needs, design authentic tasks, and sustain engagement. This exploratory study examines how artificial intelligence (AI) can serve as a teacher-facing pedagogical resource for exploring professional situations relevant to civil engineering and designing contextualized, meaningful task-based learning activities. The study involved 48 fourth-year civil engineering students in an 11-week ESP course that incorporated both textbook-based activities and teacher-mediated, AI-supported task-based activities. Using an exploratory, retrospective comparative design, the study examined students’ reported engagement across behavioral, cognitive, emotional, and social dimensions, as well as their perceptions of sustainability awareness and professional responsibility in sustainable practice. Data were collected through a study-specific questionnaire, classroom observations, focus-group discussions, and student artifacts. The questionnaire elicited students’ retrospective perceptions of both activity types at the end of the course. Descriptive findings showed higher reported engagement for AI-supported task-based activities in the behavioral, emotional, and social dimensions, while cognitive engagement received similar ratings across both instructional approaches. Qualitative findings highlighted different forms of cognitive engagement across the activities. Sustainability-related findings were mixed, with higher ratings for most items in the AI-supported activities. The largest difference concerned perceived responsibility for sustainability in professional practice, which focus-group accounts did not directly link to the activities. The findings do not establish causal effects or isolate AI’s contribution from other instructional design features. Nevertheless, the study illustrates a practical approach to integrating AI into teacher-mediated task-based ESP design and contributes to emerging research on AI-supported pedagogy in ESP.