With your permission, this site uses Google Analytics cookies to count visits and see which links readers open. You can change your choice in the footer. How Google uses this information.
This area examines AI and analytics in accounting, business decisions, and professional education.
AI, analytics & professional judgment
My research and teaching examine how AI and analytics can inform accounting decisions, how professionals evaluate their outputs, and how students build the judgment to use these tools. Related work considers accounting education, machine learning, and governance.
The IUP Journal of Accounting Research & Audit Practices, 25(1), 301-316
The paper explores how Microsoft Excel-based instructions can enhance learning in undergraduate accounting through Active Learning Theory and Cognitive Load Theory.
We address the ongoing shortage of CPAs, arguing that the traditional 150-hour educational requirement is a significant barrier to entry and often fails to equip students with the necessary tech skills for an AI-driven profession.
This systematic review synthesizes recent research about gamification in accounting education, where student disengagement and motivational obstacles frequently hinder learning.
In this opinion, Dr. Mulyadi argues that when a client brings an AI-generated second opinion on a draft tax return, firms should treat it as a request for changes that require evidence, not as a corrected return. He proposes a short record linking each proposed change to the draft reviewed, the supporting documents, and the professional's decision, with the qualified tax professional independently verifying the authority and recomputing any adjustment rather than asking the same AI tool to endorse its earlier answer. He also calls for clear limits on the documents the AI can access and on what a review does and does not cover.
Dr. Mulyadi was quoted in a Forbes article on workplace AI policies that could help companies prevent a crisis. He advised that policies distinguish low-risk uses, such as brainstorming, from consequential uses involving confidential data, financial reporting, hiring, customer decisions, or transactions, and make clear who remains responsible for the final decision. He noted that human review is meaningful only when the reviewer has the knowledge, time, and authority to challenge the output.
Dr. Mulyadi was featured in a Techopedia article on what businesses really pay to use AI. He advised companies to look at the total cost of getting a task done correctly, not just how much each model costs to use. For example, a company using AI to pull information from invoices should track how often it misses details, gets amounts wrong, or needs someone to fix its work.
In this opinion, Dr. Mulyadi argues that adding technology expertise alone will not solve audit committees' AI oversight problem, because committees cannot challenge management effectively without evidence. He proposes a concise, quarterly AI governance evidence pack for material use cases covering scope and accountability, material changes, validation and controls, exceptions and incidents, and assurance. Drawing on familiar accounting discipline such as materiality, exception aging, and the difference between a policy and evidence that a control actually operated, he urges keeping the pack exception-based and decision-focused so it does not become another box-checking exercise.
Dr. Mulyadi was featured as an expert in a Forbes article on a survey finding that only 10% of respondents feel in full control of the technology they use. He described the mismatch between being held responsible for decisions and lacking understanding or power over them as an internal control problem. He clarified that employees may overtrust the technology or quietly work around it, allowing errors and risks to accumulate.