Man vs. Machine: An Accountant’s Take on Best Practices in the Age of AI Part One: Bringing Back the Old School?

This is the first in a series of articles that will explore the benefits and challenges of Human vs. Artificial Intelligence (AI) and examine methods to maximize usage for both business benefit and employee well-being.

Frequent interaction with Artificial Intelligence (“AI”) has rapidly become a normal feature in the day-to-day cadence of our lives. From social activities, such as suggesting a location to meet friends, to professional obligations, such as formulating a report for a teacher or boss, AI has become a tool we increasingly rely on to help accomplish our daily tasks and activities. As this extraordinary technology continues to develop – and in some ways surpass – human cognitive capability, will we still need to rely on Human Intelligence (“HI”) or will AI simply make HI a quaint relic of the past? Will we become so reliant on the “artificial” that we’ll soon forget the “human” element that started this phenomenon, in the first place?  Is it possible for HI to atrophy to the point where we become hopelessly dependent on machines? And what does this all mean for business?

Similar to the introduction of other profound technological innovations that have significantly altered our world (the discovery of fire, the printing press, the automobile, the assembly line, the Internet) AI has presented us with great opportunities for advancement. Large Language Models (LLMs) – think ChatGPT or Llama – provide us with vast capabilities to improve content creation, data analytics, and productivity. On the flipside, this same technology also presents a slew of new risks and challenges that have the potential to compromise financial assets, intellectual property and privacy, or even destroy an entire business – that is, unless we maintain our HI, and in fact work to keep our HI on par with AI.

For most of the last 150 or so years, knowledge was associated with having information stored in your head that you could recall at a moment’s notice. Not very long ago, it was standard for most school children to memorize crucial information such as the state capitals, multiplication tables, etc. With the advent of the Internet, and powerful search engines such as Google, information has become democratized, and memorization is no longer such an important skill; you can just “Google it!”

Up to this point in our existence, HI has eloquently adapted to the tools and technologies we have at our disposal; in modern society, it is less important to “know” a fact or an issue as one merely has to know where to find it. Critical to this shift, the focus is now on the “operator” who must be able to discern if the answer is “right” or makes sense – this is where we must continue to evolve our HI to optimize the benefit of AI.

Frequent users of the Microsoft Office program, Excel, have all, at one point or another, experienced a formula error that was not discovered by the software, but was instead recognized and corrected by a person who deployed their HI, by manually testing and analyzing the spreadsheet in order to find and correct the error. Making sure to ‘sanity-check’ computer outputs, test citations, and question the reasonableness of those outcomes is still a gray matter job – not a silicon molecule one (although this may change in the next ten to twenty years).

While the focus continues to be on the great potential of AI, we don’t want to ignore the dark side. Namely, the risk that AI, with all its speed, access to information, and ability to replicate human logic and artistic creation, can also be used to deceive, mislead or even commit crimes that can profoundly compromise businesses and their employees.

In the past, with some effort and ingenuity, a fraudster could easily hide the fact that a check was not made out to the entity specified in the corporate books and records by intercepting mailings from the bank. It was less easy, but also feasible, to use copiers, white-out and typewriters to alter bank statements to hide or misrepresent certain unauthorized transactions. As a safeguard against this, it used to be standard audit practice to request that bank statements be periodically delivered (unopened of course) along with the cancelled checks, directly to outside auditors. The premise was that this would reduce the risk that an internal fraudster could intercept and alter the bank statement, dispose of the cancelled check, and keep the auditors from discovering any alterations or inconsistencies with the books and underlying defalcation schemes. This audit practice seems to have found its way to the retro thrift shop, as cancelled checks are no longer returned to the drawing company nor are images necessarily included with bank statements.

In today’s climate, with the advent of sophisticated image generation technology, financial documents must be even more closely scrutinized to ensure transactions are what they seem; remember, the entire Madoff fraud was possible because statements and account activity were simply created out of thin air – and that was nearly twenty years ago!  It was only through HI that people deduced that the results were too good to be true.

With the continued rapid advancements in digital crime (deep fakes, phishing scams, data piracy, digital banking fraud, etc.) heighted vigilance is a prerequisite for successfully operating in any modern business environment.

As this series continues, we will discuss in greater detail the training programs business leaders can deploy, and the steps internal controls and IT leaders can take to ensure their organizations are positioned to utilize AI and HI to the benefit of their businesses and employees.

Our next article will be a deep dive into how AI tactics deployed to perpetrate fraud on the acquirors of a business can be defeated by utilizing the critical thinking that only an experienced financial advisor’s HI can bring to bear…

About the Author

Randall M. Paulikens, CPA, CFE, CFF, ABV
Partner, DLA, LLC

Randall M. Paulikens, CPA/ABV/CFF/CITP, is a Partner in DLA’s Forensic, Valuation, and Litigation Support practice. With over 39 years of professional experience in accounting, forensic accounting, litigation support, tax strategy, and tax compliance, Randall creates solutions for his clients that help to mitigate risk and find answers and solutions to business challenges (litigation, merger, other).

Throughout his distinguished career, including roles as a Partner at larger firms (Friedman LLP, and WithumSmith+Brown) as well as smaller firms where he has represented clients ranging from small solo practitioners to large Fortune 50 companies.

Randall has also contributed to numerous articles and is a frequent presenter at major industry conferences.

Questions?
E: [email protected]

About DLA
Founded in 2001, DLA, LLC is a leading consultancy firm that provides a wide range of specialized services designed to optimize business operations and drive sustainable growth. The firm specializes in internal audit, risk advisory, IT advisory, regulatory compliance, and other critical areas, delivering tailored solutions that help clients address their most pressing challenges. Combined with a results-driven approach, as well as the experience, knowledge, and expertise of an established leadership team led by Big Four veterans, DLA serves both corporate clients and individuals. www.dlallc.com

The views and opinions in these articles are solely of the authors and do not necessarily reflect those of DLA, LLC. They are offered to stimulate thought and discussion and not as legal, financial, accounting, tax or other professional advice or counsel.