Peer Assessment in the Age of Artificial Intelligence: The ENGAGE Model at University

Peer formative assessment fosters self-regulation, critical thinking, and collaboration but is often underused due to time constraints, limited confidence in applying evaluation criteria, and discomfort with mutual judgment. This study presents the ENGAGE AI-Based model, an adaptation of the ENGAGE framework, integrating artificial intelligence tools to support peer feedback processes. Conducted at the University of Palermo within the Metacognitive and Cooperative Teaching course during the academic year 2024/2025, the research involved 150 pre-service teachers engaged in project-based tasks and AI-enhanced peer review activities using ChatGPT. Students compared and integrated AI-generated feedback with their own comments, developing critical interpretation and metacognitive reflection skills. Qualitative analysis of participants’ perceptions highlighted appreciation for the model’s structured approach and the supportive role of AI in learning, while also revealing the need for specific training on the ethical and pedagogical use of such tools. Overall, the model demonstrated potential in promoting assessment literacy and AI awareness, emphasizing that artificial intelligence can support - but not replace - human judgment in formative assessment.