Leveraging AI for improving student writing in MBA case analyses: The MBABot initiative

Abstract

This teaching brief discusses the development and implementation of MBABot, a custom GPT‐based AI tool designed to support writing‐intensive decision sciences course titled Delivering Business Value Through Information Systems . This MBA‐level course examines the role of business leaders in evaluating information systems and information technology to deliver business value and sustain competitive advantage. Recognizing the importance of timely, structured feedback in developing analytical reasoning and professional decision writing, the initiative addresses instructional constraints such as large class sizes and limited grading time. MBABot integrates retrieval‐augmented‐generation (RAG) with course‐specific materials to provide context‐aware feedback aligned with explicit grading criteria and structured decision tools (e.g., the DECIDE model and a weighted decision matrix). Survey‐based evidence suggests MBABot supported several debiasing strategies, particularly problem decomposition, consensus building, and motivation to improve. Students also reported limitations, including redundant or unnecessary feedback that reduced trust in the tool. The brief underscores both the promise and constraints of AI‐enhanced feedback systems in decision sciences education, emphasizing responsible implementation and pedagogical alignment.