Tepper School Accounting AI Research Lab Launches New Website

The Tepper School of Business at Carnegie Mellon University has launched the Accounting AI Research Lab (AARL), an interdisciplinary initiative designed to integrate modern computational tools—such as machine learning and graph neural networks—with the fundamental, centuries-old principles of accounting. Accounting is one of the oldest business disciplines, originating from record-keeping that dates to ancient Mesopotamia. For millennia, accounting has served as the language of commerce and the foundation of organizational accountability. Today, as artificial intelligence reshapes business and society, accounting stands at an inflection point. The AARL's core mission is to advance both accounting science and practice in the service of this ongoing profound societal transformation.

The AARL is an interdisciplinary effort structured to bridge the gap between academic accounting research and practitioners' operational challenges. Originally founded by Professor Pierre Jinghong Liang in 2023 and now co-led with Professor Gaoqing Zhang since 2025, the lab brings together three key communities: practitioners confronting real-world problems, accounting scholars with foundational theory, and computer scientists developing computational methods. The lab’s many projects adhere to three guiding principles: address genuine problems from practice with significant impact potential; leverage foundational accounting theory; and deploy computational tools responsibly.

In practice, AARL researchers are already producing impactful work. For instance, in partnership with industry sponsors, they have developed graph-based anomaly detection systems that leverage the mathematical structure of double-entry bookkeeping. By representing accounting records as complex networks, these systems can identify irregular patterns at scales impossible for traditional methods. The recent Accounting Classification Entropy project applies Claude Shannon's information theory framework to measure the informational richness embedded in financial statement classifications. By calculating the entropy of accounting line items and their interrelationships, this new toolset moves beyond single-dimensional metrics (like earnings-per-share) to objectively measure the information conveyed by the entire structural dimension of classified financial statement numbers.

The Tepper School of Business has a long history of shaping accounting scholarship by applying rigorous analytical methods to management problems. This legacy began with William Cooper, the first modern accounting academic, who pioneered a scientific and interdisciplinary approach to the field. His doctoral student, Yuji Ijiri, became a preeminent theorist of measurement, famously developing triple-entry bookkeeping. The tradition continues with figures like Shyam Sunder, a CMU Ph.D. alumnus whose groundbreaking work spans financial reporting, information economics, and experimental markets.

The AARL is funded by Quantinno Capital Management LP and has received past support from Carnegie Mellon, the Tepper School, and other grants.