Machine Auditing of Financial Statements
In this months publication of CFO Magazine one article asks the question –
“Can we trust the auditing of financial statements to artificial intelligence?”
The articles contributors Bill Brennan, Michael Baccala, and Mike Flynn pose a very real question that many in the office or finance are faced with today. The prolific growth of enterprise information, combined with the emerging nature of distributed enterprise organizations has dramatically added to the Office of Finance’s burdens. Tools like IBM’s TM1 and Planning Analytics, Jedox, and Qlik Sense have helped finance managers keep up with the pace of change, and get ahead of the curve using multi-dimensional analysis, what-if scenario modeling, and rolling forecasts. All of these capabilities combined have helped to transform the Office of Finance from number crunchers to partners in the performance of the business by allowing for more in-depth analysis.
We capture and consolidate everything directly using Planning Analytics in the Cloud which saves around 32 hours per month in forecast preparation time and there is no risk of manual error.” –Financial Planning Manager, Hunter Industries
The benefits of advanced analytics to Finance can be both quantitative and qualitative. Hunter Industries saves time in forecast preparations – quantitative. They take that saving and put it into improving analysis and insight that has lead to improvements in operations – qualitative. Those benefits aren’t always discernible at the outset. I suspect the benefits of trusting the auditing of financial statements to artificial intelligence aren’t either.
The CFO article states that “visionaries foresee the day when AI will enable auditing that is a continuous and real-time process, not a prolonged exercise requiring large teams of accountants working overtime.” This begs the question – What will the Office of Finance replace these extended activities with? Is Finance being replaced by Watson like intelligence? The answer is, it depends. Finance leadership looking purely to control costs will undoubtedly be tempted to use these tools to do so. On the other, Finance leaders seeking to control costs and augment their analysis capability with enhanced machine intelligence will out compete the cost controllers.
The answer to the question “Can we trust AI to audit,” depends on how much faith can be put into the dataset. How is financial data acquired? Is there access to raw business data? Can it be checked for accuracy and alignment? Ultimately automating some of these finance process is beneficial, if done right. However, doing it right takes experience. Organizations undertaking these projects should seek to answer the “So That…” question as well.
- So that… We can decrease costs.
- So that… We can increase market agility.
- So that… We can improve operations efficiency.
- So that… We can spend more time in analysis and reporting.
Whatever the drivers are, define those clearly first and foremost. Then seek out others who had done it before and learned the hard lessons. If you’re considering any planning analytics for your organization undergo a benchmarking and readiness assessment, and check out how Hunter Industries is improving their own Financial Planning and Analysis now, and maybe someday using Augmented Intelligence to make more improvements.