Abstract

The rapid emergence of artificial intelligence (AI) tools has risen new opportunities to support students in developing evaluative judgment and feedback literacy in academic writing. However, learners may find AI generated feedback misaligned with assessment criteria, lacking practical relevance, or uncomfortable given its non-human source.

This paper examines the use of artificial intelligence (AI) to support the development of critical thinking and peer-feedback literacy in Academic Writing course among undergraduate IT students. The intervention employs a student-centered approach in which students first generate peer feedback independently, then interact with AI tools, and finally critically evaluate AI-generated suggestions before producing a revised response. It is argued that this structured critical engagement with AI can enhance the quality of peer-feedback and improve students’ academic writing performance. Using a classroom-based quasi-experimental design, the study analyzes changes in feedback quality and writing outcomes before and after the AI-supported intervention. Data sources include student writing samples, peer-feedback forms, and reflective responses. In contrast to concerns that AI-generated feedback may be misaligned with assessment criteria or lack practical relevance, this study suggests that guided interaction with AI can promote greater feedback specificity, encourage higher-order thinking, and strengthen students’ peer-feedback skills.

The findings are expected to demonstrate that integrating AI through a critical, human-centered framework provides a practical and pedagogically sound model for enhancing feedback practices in higher education. The study acknowledges several limitations, including the relatively small, single-institution sample (40 undergraduates students of Astana IT University, Kazakhstan), one assignment-based experiment. The paper potentially contributes to ongoing discussions on the effective integration of AI in feedback instruction practices and the development of students’ capacity to evaluate and question AI-generated outputs.