Essay by: Dr. Jennifer A. Miller, CDFM-A, Washington Chapter

Artificial Intelligence is growing, fast! We are currently witnessing the expansion of artificial intelligence or “AI,” commonly described as systems with intellectual processes characteristic of humans, such as the ability to reason, discover meaning, generalize, or learn from experience. While we have embraced related capabilities like robotic process automation for a few decades, AI represents a distinct paradigm shift. In the following, I will offer some of the many roles I think AI should serve in the future of financial management. My offering will include examples of potential applications and their potential impacts.

First, there is AI in the form of a variational autoencoder (VAE). I promise it is not the latest automobile part thieves are making headlines for stealing. However, it is a form of AI supporting the claims AI is a way to allow humans to experience more of the essence of work and less of the weighty load. VAE is a type of neural network that learns to reproduce its input, and also map data to latent space. In other words, VAE can create samples by first sampling from the latent space. These VAE latent spaces are intentionally continuous, allowing for easy random sampling and interpolation. Auditing is an existing, reliable use of VAE, which makes it one of my top choices to serve in the future of financial management. VAE neural networks should be used to learn a representation of journal entries able to provide a comprehensive sampling to auditors. Thus, VAE can be a significant load lifter in financial statement preparation and audits. The Department of Defense is decades into challenged progress when it comes to financial improvement and audit readiness, undergoing audits, and achieving opinions, let alone the more favorable types of audit opinions and reports. Headlines and various reports capture the intangible and tangible weighty costs too. This is where I see VAE’s role. VAE has application in the intensive sampling process of these high volume and sweeping audit sample efforts year after year. The sooner VAE is employed, the sooner the return on investment while lightening loads for those in our often-overburdened areas like the Defense Contract Audit Agency, Defense Contract Management Agency, and many accountants, auditors, internal reviewers, and others sleuthing through the sampling sludge start to finish. The professionals of these agencies, series, and others are then able to fill the human gaps AI cannot, such as creative problem solving, intuitive solution seeking, mental model making, and emotional leveraging. This is one example of potential applications and impacts, the next is unsupervised.

Next, we will turn to a more popular natural language processing AI: Generative Pre-trained Transformer (GPT). GPT options are under the Generative AI (GAI) umbrella, and GAI is another version of AI I think should serve in the future of financial management. GAI is what we are seeing generating text, images, audio, and other data by way of generative models frequently in response to prompts. You have probably seen, heard, and read about the headline-making new content and capability GAI offers like ideas, conversations, stories, term papers, speeches, and more. One of the most popular GAIs is ChatGPT while GovGPT may be more popular for those interested in and around the US local and state government. This leads me to believe AI in the form of GAI has a tremendous future role in financial management at the local, state, but also federal levels of the US government and beyond. I have seen some live examples of GovGPT as compared to ChatGPT – both are superior options in almost all subjective and objective survey categories like speed, accuracy, quality, and more. At this time, local and state governments use GovGPT to create a unified platform for communications with input and outreach while saving time, money, and effort for automatic discourse. GovGPT exists to assist stakeholders and glean insights from public comments, then process feedback for concerns, trends, and data-driven decision support. The turnaround time between writing officials and organizations until seeing a response results in a more responsive, timely, and transparent government for the people, by the people. Such impacts at the federal government level, and in the financial management area, would be welcomed by many. It seems logical to believe this could be a great next application in the federal financial management space to reach the same types of positive gains but in monstrously larger measures for mighty stakeholder impact – taxpayers, voters, congress, etc.

Finally, a third role I think AI should serve in the future of financial management is Generative Adversarial Networks (GANs). GANs work by having a generator of content exchange with a discriminator until the outcome is sufficient to a real product that the discriminator cannot reliably tell the difference. One ripe application for GANs in financial management is financial fraud detection. The Bureau of Justice Statistics describes financial fraud as, “The intentional misrepresentation of information or identity to deceive others, the unlawful use of a credit or debit card or ATM, or the use of electronic means to transmit deceptive information, in order to obtain money or other things of value. Fraud may be committed by someone inside or outside the business.” These days financial fraud scams are at an all-time high while crypto currencies confuse the population too. Financial fraud detection could apply and yield important impacts in areas like the Internal Revenue Service’s (IRS) annual Dirty Dozen list. This list helps readers to be aware and armed against tax scams affecting taxpayers, businesses, and professionals. Financial fraud may include credit claims, deductions, fake charities, offshore accounts, evasion, conservation easements, micro-captive insurance, misused trusts, and much more. Potential impacts of using GANs supporting financial fraud detection include prevention of victims, fraud and theft cleanup, drained finances, stigma, and other negative outcomes on top of emotional pain and suffering. However, the application of financial fraud detection in financial management does not stop at prevention since the IRS’ Criminal Investigation Fiscal Year 2022 annual report disclosed the organization initiated more than 2,550 investigations and achieved a 90% conviction rate with enforcement actions focused on tax fraud, money laundering, and cybercrimes. This worldwide effort identified over $31 billion from tax and financial crimes. This is amazing and all the more support to continue pursuing financial fraud detection with the best tools possible to surpass prior years of investigations, convictions, and collections for even greater impacts!

In closing, we are actively watching and participating in the expansion of AI worldwide. We are familiar and more comfortable with related tools while AI is distinctly different. Nonetheless, AI is a tool with many potential roles to play in the future of financial management. I have shared my top three roles for AI plus examples of existing and possible applications followed by their potential impacts. Let us keep our AI eye on the prize!

This essay was awarded the Distinguished Honor for the Essay Contest.

See all 2024 SDFM Award Honorees.

About the Author:

Dr. Jennifer A. Miller is the Business Operations Manager for the Defense Health Agency’s Chief of Staff at the Defense Health Headquarters in Falls Church, VA. Previously, she served as a Financial Manager for the Defense Health Agency’s Reserve Health Readiness Program. Over the past 14 years, she’s supported the Defense Department’s fourth estate, and the Air Force and Army’s active and guard components along the east coast in capacities including financial management, data and information management, budget, analysis, cost, audit, and accounting. She is a certified project management professional, certified government financial manager, a certified defense financial manager with acquisition specialty; and a member of the American Society of Military Comptroller’s Washington Chapter, the Association of Government Accountants Northern Virginia Chapter, and AFCEA.