Discussant's response to "AUDITOR'S ASSISTANT: A knowledge engineering tool for audit decisions";

              Discussant's Response to "AUDITOR'S ASSISTANT: A Knowledge Engineering Tool for Audit Decisions"
John B. Sullivan
Deloitte Haskins & Sells
There is much about this article that I like. The focus of this article is on the development of an expert system designed to help the field auditor make more efficient decisions about the level of audit testing required. Both these topics are high on the list of priorities of all of the national accounting firms. A significant portion of the budgets of most major firms is being directed toward the development of personal computer-based expert systems to increase audit efficiency.
The authors recognize the need for the active interaction of the field auditor in the use of an expert system. Too many articles fail to give proper credit to the level of knowledge which resides in the audit engagement team. Firms attempt to keep turnover at the partner, manager and senior levels to a minimum. As a result, audit engagements are frequently staffed by an engagement team with ten or eleven years of total client experience. I believe this paper attempts to give proper recognition to the benefits to be gained by tapping into that experience. The authors stress the importance of understand­ing
the client's business environment. In our firm's approach, we list "Understanding the Business" as the first step of the business review. Although this differs slightly from the authors' "understanding the client's business environment," I believe they both recognize this area as the first and most important step in an audit.
The article is also one of the few which focus on the audit team and the decisions which must be made by that audit team. We need more articles on and research in this important area of audit practice.
Unfortunately, I believe that my negative comments outweigh my positive comments. In summary, I do not believe that the authors are on the right track yet. My general impression of the knowledge engineering tool described in the paper is that it will not be widely endorsed in practice and it may be flawed in theory.
In my view, some of the fatal theoretical flaws of the tool involve a failure to appreciate the complexity of the audit process. Individual pieces of audit evidence frequently impact more than one assertion, and indeed may impact other areas or the entire audit. For example, one piece of audit evidence which indicts the integrity of management may lead either to a qualification or a disclaimer of opinion.
In addition, I do not believe that the authors' system, as described in this paper, gives enough weight to the importance and interaction of individual audit procedures. In Scenario One, the authors indicate revision of the nar from 0.83
80