Undergraduate international students’ perceptions of AI-generated feedback: A mixed-methods study at a Canadian university

Abstract

The rapid adoption of Artificial Intelligence (AI) has sparked debates on its benefits and risks, yet little is known about its impact on trust and role perception in higher education. This study examined how AI-generated feedback influences students’ trust in instructors and their perceptions of academic integrity. Building on existing research on AI-human interactions, we explored whether students trust AI-generated assessments more or less than human feedback and how this dynamic affects their learning experience. 

Using a mixed-methods approach, this study analyzed student attitudes toward AI in academic settings. The first phase employed a survey to capture broad student perceptions of AI-assisted feedback, while the second phase involved in-depth interviews to explore themes of trust, ethics, and emotional responses. Reflexive thematic analysis (Braun & Clarke, 2021) was applied to identify patterns in students’ experiences, highlighting concerns over fairness, depersonalization, and ethical considerations. Participants identified personalized, meaningful feedback as a core instructor responsibility that AI cannot fully replicate. Participants see value in using AI as a supportive tool when its outputs are transparent, guided by educator expertise, and clearly communicated. 

Findings underscore the need for structured institutional policies that ensure AI complements rather than replaces human engagement. This paper will explore how universities can develop ethical guidelines that balance AI’s efficiency with the critical role of instructor-led mentorship, fostering sustainable and student-centred learning environments.