Generative Artificial Intelligence Student Feedback in Large University Accounting Classes: Prompting Techniques

Abstract

The increasing student enrolments in universities accounting programmes has exacerbated the difficulty of providing timely, personalised and pedagogically meaningful feedback to student. Generative artificial intelligence (GenAI) tools including ChatGPT offer new opportunities to automate and scale student feedback processes, however, questions remain regarding their educational validity, contextual relevance and disciplinary reliability. This study reviews literature on how structured GenAI prompting frameworks can improve the quality, accuracy and ethical grounding of AI-generated feedback in large accounting classes. Using PRISMA 2020 guideline, 26 reviewed articles were extracted from Scopus and ScienceDirect. The study proposes the CLEAR-RAP guided GenAI feedback models to ensure that feedback remains rubric-aligned, context-sensitive and pedagogically coherent. Based on review a GenAI prompting template is also proposed to operationalise this framework for classroom application, supporting transparent, personalise and timely feedback that complements teacher judgement. The findings underscore that well-designed prompt can address the gap between AI automation and instructional intent, supporting ethical and effective human-AI collaboration in assessment. This study advances prompting engineering as pedagogical methodology by offering a replicable framework and template adaptable across disciplines.