There are Critical Precursors to AI, Including Verified Financial Intelligence
Part Two in a Series Addressing Advances in Forensic Accounting and Financial Forensics
The author of this series of articles is co-founder of Valid8 Financial Forensics Software, a Seattle, WA and Boulder, CO-based professional service provider. This second article, and those in the forthcoming series, set forth techniques used to investigate financial fraud allegations and the reliance and value of artificial intelligence in these investigations.
My first article in this series made the case that changing financial markets were setting the stage for an increase in financial fraud cases in the years ahead. As investigations increase, so will the need for technologies that enhance the ability to gather evidence, analyze data, and speed up the ability to make clear determinations. While artificial intelligence’s (AI) impact will likely be significant and is garnering the lion’s share of headlines, AI alone will not fix the age-old problem of garbage in/garbage out. In fact, AI highlights one of the key truisms of forensic accounting, investigations, and valuation—all professional opinions need to be based on evidence, meaning all data must be verified, not just supplied.
Today’s precursor technologies can help ensure that AI draws upon verified data when deployed. Verified Financial Intelligence (VFI) platforms are emerging as foundational in front of AI applications. For those unfamiliar, VFI is a SOC2 enterprise technology that extracts financial data, builds integrated data models, and leverages algorithms to verify evidence; then traces an accounting cycle, ultimately providing a visualization of an entire revenue cycle. This software category goes beyond client-supplied data aggregation, collecting evidence from multiple sources—including client accounting systems, bank statements, brokerage statements, check images, wire details, and transaction lists. Algorithms then compare evidence with accounting records and trace the flow of funds between accounts and legal entities, resulting in comprehensive, verified evidence that can be relied upon in a court of law. It is this data set—verified financial data—that is then run through an AI application. Jumping straight to an AI application, without any level of data verification, risks losing the benefit of AI. Frankly, it is a critical step in our work regardless, as many issues arise from incomplete or improper verification procedures.
As an example, a recent case in which I was involved showcases how significant VFI platforms can be used in speeding the time to opinion. In this situation, a court judgment was entered against a high-profile trial lawyer. The court ordered him to pay the partners from a prior firm. The trial lawyer claimed he did not have the money and the plaintiff suspected he was hiding it. Scott Sims of Frank Sims Stolper, a California-based national law firm, was representing the plaintiff and filed subpoenas for relevant banking activity from 35 different accounts. Upon receipt, he was up for hours at night looking through the data to see patterns and areas of concern. He started to enter the transaction data into a spreadsheet but became overwhelmed with the sheer amount of data. He also did not have the budget to outsource this work. My firm was engaged to work on the assignment and we began our engagement by uploading more than 600 documents into a VFI platform. With the power of the software behind us, we reconciled statement data within hours. And within minutes, the software’s algorithms identified money suspiciously being moved between accounts. Within days, the software reviewed 2,633 transfers totaling $144 million and identified multiple million dollars of stolen funds. In addition to accelerated analysis and fraud identification, a fully reconciled and comprehensive set of transaction data was ready for analysis in less than 24 hours. Most important of all, Scott was able to deliver a clear, evidence-based opinion on the case.
The ability to aggregate data from multiple sources, accelerate analysis, identify fraud, and create a defensible data set and a clear case are critical to the forensic profession. It is an enormous step forward to do so in a way that leverages proven technology and is mindful of human capital and available billable hours.
For those of us who have worked on cases involving data at scale, technologies that generate verified financial evidence allow us (for the first time ever) to leverage technology during the initial evidence gathering and data modeling; enabling us to truly “follow the money.” Too many investigations leverage sampling and other techniques, but today’s technology choices make that decision unnecessary. VFI platforms allows for every piece of evidence collected to be analyzed and a powerful data visualization model to be produced. With the emergence of AI and the arrival of foundational technologies, such as VFI, it is evident that “the times they are a-changin’.”
Tod McDonald, CPA, CIRA, was the lead on an investigation that unraveled a $150 million dollar agri-business Ponzi scheme in Washington. He was a senior auditor at Ernst & Young early in his career and has spent decades working to clean up and turn around complex accounting and financial situations. Today, he is the co-founder of Valid8 Financial, helping forensic accountants, fiduciaries, attorneys, investigators, and auditors eliminate data prep work associated with finding and analyzing evidence of financial records.
Mr. McDonald can be contacted at (206) 920-1144 or by e-mail to: linkedin.com/in/todmcdonald/ or t.mcdonald@valid8financial.com.
