5 leading technology-driven techniques used to investigate fraud
How tech and analytics drive efficiency and provide deeper insight
INSIGHT ARTICLE |
Today’s fraud investigators have a variety of tools and techniques at their disposal to deliver meaningful analysis and insights at ever-increasing speed. The following five tools are among the most influential and often used to advance investigations while minimizing cost:
- Artificial intelligence (AI)-driven data processing technology
- Document review technology
- Automated online data research and extraction applications
- Analytic libraries
- Data visualization
1. Reduce lead time and fees through AI-driven data processing tools
Investigations often necessitate the review of bank or credit card statements. The scope of an investigation can range over several years and include review of multiple bank accounts and respective statements. Historically, reviewing bank statements involved a significant manual effort—and considerable time and money.
Investigators today utilize AI-driven technology to convert numerous bank and credit card statements into analyzable datasets in a matter of hours. In some cases, this technology is bolstered by additional timesaving features such as automated transfer matching and cash flow diagramming, allowing the investigator and clients to quickly assess transactional volumes and account relationships and effectively focus investigative efforts.
Case study: RSM was engaged to perform forensic financial analysis in a high net worth estate dispute involving over 40 individual accounts spanning a 15-year time period. RSM’s AI-driven technology extracted the relevant data from the statements in an analyzable format and identified all transfers between the 40 accounts, allowing our team to quickly understand the flow of funds. This tool provided RSM with the means to easily isolate and further analyze fund transfers that went outside one of the known individual accounts. Our team arrived at actionable results in a matter of days versus weeks and passed along tens of thousands of dollars in estimated cost savings to our client.
2. Find the needle in the haystack more easily with document review technology
Efficient digital forensic data collection and analysis is a key component to many investigations. Investigators are often required to review emails and data stored in cloud servers or on a physical device such as a laptop or mobile phone. In today’s data-saturated world, this process typically involves identifying relevant information among hundreds of thousands of files.
Smart investigators leverage significant advances in document review technology, such as natural language processing, computer vision and sentiment analysis, to effectively distill voluminous document productions and guide future analysis. This AI- and machine-learning approach helps investigators rapidly assess, refine and re-scope investigations and complements more traditional approaches such as targeted keyword searching.
Case study: RSM led an investigation initiated by a financial institution on suspicion of residential mortgage loan fraud involving a loan officer employed by the financial institution and a real estate agent. A forensic image of the loan officer’s laptop was created. Computer-vision and machine-learning automated document classification technology helped RSM quickly identify files with relevant information, including falsified pay stubs, W-2s, employee verifications and bank statements. Investigation findings and conclusions were presented to the financial institution’s audit committee for further investigative purposes.
3. Enrich analysis and uncover insights through connectivity to external datasets
One of the most common techniques in asset misappropriation and procurement fraud involves utilizing shell companies or businesses associated with a fraudster’s family or known associates to siphon funds from an organization. Therefore, understanding an investigation subject’s relationships is an essential element of many financial investigations.
In the past, investigators could spend days performing research and extracting relevant information from background reports and other databases. Luckily, today’s investigators have the ability to automate much of this online data research and extraction in order to efficiently examine identified individuals. With a few identifying data points for each individual or corporate entity, a database can be created to uncover relatives, associates, affiliated businesses, addresses and phone numbers. Data analytics routines can then compare these results to a list of key vendors, employees, partnerships, advisors and more to identify potential areas of risk such as conflicts of interest and other relationship risk areas.
Case study: RSM was engaged by a private equity fund to investigate suspicions of self-dealing at a portfolio company. The client provided RSM with a list of 35 employees, which our team subjected to our proprietary automated conflict of interest screening application. This automated, efficient process identified multiple shell companies, consulting firms and individuals related to executives of the subsidiary that were previously unknown to the private equity investor. These results aided the investor in forthcoming buyout negotiations with the subsidiary.
4. Leverage previous experience through the development of analytic libraries
Every investigation has different players and fact patterns, but over the years investigators have identified various commonalities and indicators that can aid in systematically assessing risk and prioritizing detailed analysis. Prior to the use of advanced data analytic tools, investigators would draw upon previous experience to design a discrete set of analytics for each investigation.
Today’s data analytics professionals have eliminated much of this lead time by building dynamic and repeatable routines, using leading technologies such as Alteryx, Python and R to rapidly analyze books and records for indications of fraud, waste, abuse and corruption. Such tools can also help identify areas of process improvement to increase the efficiency and accuracy within a client’s financial reporting processes. Libraries of hundreds of tests can be built to quickly draw upon and tailor to each investigation.
Case study: RSM was engaged to analyze five years of purchasing and disbursements data covering nearly $70 billion in payments from two separate SAP systems for a large international company. We leveraged our existing deep library of SAP-specific analytics to rapidly identify issues warranting further investigation, such as non-purchase order (PO) spend, PO line items with a large number or value of price changes, duplicate payments, disbursements well in advance of payment terms (including those resulting from changes to SAP’s invoice document baseline date), and vendors with short-life bank accounts, among other items. We then trained select internal audit personnel on how to understand, interpret and communicate the results of these analytics on a go-forward basis.
5. Using data visualization as a catalyst for collaboration
Often companies are overburdened with several ongoing investigations, which can be difficult to track and manage. To enhance the company’s understanding of risk and ability to respond to potential investigations, it should track key performance indicators such as response time and how allegations are being reported. Read our article for more information on how organizations use analytics to address the heightened risks associated with the recent surge in whistleblower tips.
Clear, powerful, interactive dashboards and diagrams foster collaboration and generate insights. Investigators utilize leading data visualization software and techniques to unpack complex concepts and processes into understandable and practical deliverables.
Case study: A global consumer products company brought RSM in to assist with improving its global ethics and compliance program. Our team created a standardized dashboard that provides insight into global case activity, including various illustrative charts and dynamic maps. This dashboard allowed the head of investigations to quickly and easily provide updates to senior management including near-real-time investigation updates and key performance indicators without requiring manual data extraction or report preparation.
Technology evolves quickly, and leading, forward-looking investigators are leveraging innovation and analytics tools to streamline and enhance key processes. By working with an investigator that is well-versed in emerging tools and tactics, you can gain more insight into potential issues and accuracy within critical data while also cutting costs due to inherent boosts in efficiency. To learn more about an effective forensic data analytics toolkit, download our infographic.