Artificial Intelligence for Audit, Forensic Accounting, and Valuation

CHAPTER 6Building an IAA Audit Firm: The Planning Toolkit

IN THE PREVIOUS TWO CHAPTERS we observed the power of machine learning. We saw that machine learning offers a host of methods, approaches, and models to develop powerful solutions. We also recognized in Chapter 4 that machine learning is only one of the four tools of automation for audit. The other tools are robotic process automation (RPA), process mining, and expert systems. In this chapter we build the basic design and analysis toolkit that will enable us to know how to develop intelligent audit automation (IAA) solutions and it will also give us an understanding of intelligent agents.

WHY PLAN AND DESIGN?

Imagine the alternative to not having a structured process for performing design activities. Recall the times when expert systems were presented as the greatest thing since sliced bread.

Expert systems were viewed as the ultimate solution to a wide spectrum of accounting problems (Meservy et al., 1992). Their eventual decline reveals our inherent inability to predict the future. In some ways, we paint the future in terms of what we would like it to be rather than what it can be. In a journal article, Alles and Gray (2016) touch upon an important historical lesson. Referring to the Ellis report from 1993, they remind us regarding our euphoric optimism about new technologies. The Ellis report predicted that by 2013, audit fees would swell from $14 billion to $21 billion. Just the e‐commerce assurance part was estimated ...