Although the field of accounting sometimes is seen as boring, the reality is that it is often on the cutting edge of new technological developments.
For example, the personal computer revolution didn’t really pick up steam until the invention of the spreadsheet. Digital spreadsheets were such a revolutionary time saver for accountants that many were willing to plunk down their own money for expensive PCs and bring them into the office since the corporate mainframes couldn’t compare.
Now, the field of artificial intelligence is the next big thing in technology. Just as they did with PCs, some accountants are already shelling out for their own ChatGPT accounts to try out the incredible reasoning and natural language skills it has shown.
The role of artificial intelligence in accounting will be hammered out by those individuals as much as by big business deals negotiated in luxurious boardrooms. Coming from both angles, the impact of artificial intelligence on the accounting industry is sure to be big.
AI in Accounting and Finance Is on the Way
Developments in AI have mostly been happening in the far off crystal castles of computer science research facilities. But the release of ChatGPT in 2023 brought some of the incredible breakthroughs of the technology into offices and living rooms around the country.
Plenty of accountants have been trying it out as well, not to mention college students of all stripes. Even lowly chatbots have been making waves in academic circles as students use them for homework and professors try them out for grading papers.
While feeding a common math problem into today’s chatbots is an iffy proposition, it’s pretty clear that their ability to manipulate data, spot long-term trends, and quickly answer complex questions will have a big impact on business in general, and on accounting in particular.
With the potential to replace many low-level accounting jobs within years, no one studying or working in the field today can afford to turn a blind eye to the evolving field of AI.
Exploring Potential Advantages of Artificial Intelligence in Accounting
Artificial intelligence describes programs that can take on some of the logical and reasoning abilities of humans.
While computers have long been able to outperform humans in feats of math and calculation, those operations are all clearly defined and linear. A calculator can easily perform long division with huge numbers, and a spreadsheet can perform many parallel calculations that would take a person a lifetime to complete.
AI brings new capabilities in language and reasoning to computer systems.
But neither one of them have any ability to connect those numbers to concepts like receivables, or explain the impact of them to a client. Computers run programs, and programs can only deal with what they have been coded to deal with. Problem-solving and flexibility aren’t part of the package.
AI changes that. With new techniques and technologies, it gives computers some limited ways to think through problems like people. And since they think much faster, they can come up with novel and insightful solutions faster and more efficiently.
Understanding AI in Accounting
Accounting professionals are no strangers to computers. In many cases, they will already be familiar with the insights and reporting tools offered by data science, a field that is extensively powered by a technique from AI called machine learning.
So it won’t come as any surprise that this new computational tool will also have some major impacts in the world of accounting.
But looking at the weird conversations generated by ChatGPT or the bizarre images floating up from DALL-E and other generative AI tools, it might not be immediately obvious what exactly the benefits of artificial intelligence in accounting will be.
Let’s take a look.
Process Automation Can Increase Efficiency to Benefit AI in Accounting
Although math gets most of the attention when people think about accounting, in reality it’s a lot of paper shuffling. Only the paper today is virtual, and the shuffling all happens in the minds of machines.
That’s a clear advantage for a thinking machine, and one of the big ways that AI is likely to impact accounting.
It’s called process automation, and it’s already important in most accounting departments. Big accounting software packages automatically stuff numbers from electronic orders into invoicing; they match up deposits and reconcile accounts, file expense reports, and tumble all of that data together into financial reports that go out to clients or executives.
Each of those steps would have to have been performed manually by bookkeepers or accountants in the past. But today those individuals take on more of an oversight role. They make sure the process runs smoothly rather than running it themselves, noticing discrepancies or interpreting results.
AI has the potential to take on that oversight role, too. With similar reasoning skills, it can understand when the numbers, metaphorically, don’t add up even when the math is right. It can sense trends and explain results.
Essentially, AI in process automation can take on the whole role from top to bottom.
Artificial Intelligence to Detect Fraudulent Account Activity Is Already Changing Financial Security
Financial controls are an important part of accounting work. Keeping humans in the loop has been important because spreadsheets don’t know when they are being lied to. A person can sense when something is off and spot unusual patterns of activity.
But that’s something that AI can also do, and it can do it faster and better. Machine learning algorithms are already a big part of fraud detection software systems at big banks and credit card companies. With an ability to take in volumes of data far larger than a human can look at, they have a more sensitive nose for trends. Even transactions that are subtly outside the norm may draw their attention.
They also work in the other direction, reducing false positives through more accurate evaluation. Algorithms that learn over time can figure out when families are on vacation for spring break and spot the sequence of unusual, but normal, transactions that come along with it, for example.
All of these capabilities can roll over into auditing work in general. In fact, big firms like PwC are already using AI in audits. The statistical systems behind their machine learning algorithms are faster and more effective than an army of auditors going through the books by hand.
Because AI doesn’t sleep or get bored, it’s also enabling new tools for auditors like continuous auditing. Rather than waiting until weeks or months after transactions are closed, it can go through the books nearly in real-time, flagging problems for instant resolution.
Using AI in Accounting Customer Service
Maybe the clearest use of generative AI in accounting will come in the role of customer service and communication. While ChatGPT and similar chatbots have proven to be spectacularly bad at math, they are tremendously fluent in English… and other languages.
Used as translators, onboarding tools, and a first stop for answering phones and online inquiries, these tools can take much of the load off accounting teams. With ties into the data backend of the customer accounts, they may be able to generate instant reports or answer in-depth accounting questions immediately and without human intervention.
Big service providers like SAP and Salesforce are already integrating generative AI copilots into their systems. Accountants will find new capacities appearing in cloud-based accounting platforms almost overnight.
Artificial Intelligence in Accounting and Finance Research
Although numbers and math are all pretty static concepts, the rules of accounting are constantly changing. GAAP, the Generally Accepted Accounting Principles, and the tax code of almost every federal, state, and local jurisdiction are constantly evolving.
Add on to that rule changes from other financial regulatory bodies like the SEC (Securities and Exchange Commission) and the changing needs of clients, and it’s easy to see why it takes many hours of continuing education each year to retain a CPA license.
AI tools are stars when it comes to taking in, absorbing, and summarizing large amounts of information, however. Dropping a generative chatbot on top of a mountain of new updates to the tax code could allow accountants to ask simple questions about changes and get instant answers, without digging through piles of dense regulatory language.
Even better, AI may summarize all important changes, based on knowledge of a firm’s past practices and needs. It may highlight critical information that otherwise wouldn’t come to light until late in the process.
Looking At the Advantages and Disadvantages of Artificial Intelligence in Accounting
Not everything involving the role of AI in accounting is sunshine and roses.
For starters, current generations of generative AI are terrible at math. There’s no easy fix for this; it’s in the statistical nature of the models to have varying answers based on their training data. Unfortunately, the answer to 2+2 isn’t variable. While not all accounting involves math, accuracy in calculation is very much a baseline expectation.
Accountants may not have to worry about their jobs as long as AI continues to show the math skills of a distracted eight-year-old.
Also, the need for accuracy is paramount when it comes to accounting advice. The regulations on financial and accounting device are strict and have real teeth. Inaccurate advice to clients comes with consequences. And once again, these are not areas where a statistical response is appropriate.
The state of the art advances quickly in this field, though, so it’s hard to say what solutions tomorrow may bring. But there are still some serious security and ethics considerations accountants will have to make when thinking about AI tools.
Privacy and Security Disadvantages of AI in Accounting
AI generally doesn’t store sensitive information directly. But some generative AI tools have shown a disturbing tendency to cough up bits and pieces of their training data verbatim. If that data happens to be personal financial data, it’s going to be a bad day for the company responsible.
Even if it’s not storing sensitive data, AI will have to be given access to that information in order to work its magic. While accountants are already familiar with standard security controls to lock down data to those who need to know, the way AI works can leave it vulnerable to deception.
In the same way that a human accountant may be duped by a phishing email into releasing sensitive data, an AI can be manipulated by skilled prompt engineers to show information that it was supposed to keep private. Red team experts have caused ChatGPT and Amazon AI agents to cough up revealing private data of their employees simply by asking the system to repeat a word endlessly. Similar vulnerabilities could cough up tax data, banking information, and financial status unless new protections can be devised.
Considering the Ethics of Using AI in Accounting
For the most part, discussions of ethics in AI revolve around ways that machine learning tools might be used to deceive people or misappropriate personal information. In some senses, those concerns don’t apply as much in the accounting uses of AI. The math will always have to add up, and the data is all being used to perform the same functions already today.
Accounting comes with some unique ethical considerations when it comes to uses of AI.
Accountants have strict ethics rules that have come to the field through lessons learned over the years. Transparency and accountability are important principles that allow work to be checked.
Yet the way that machine learning systems build neural networks today is anything but transparent. The many layers of weighted neural connections that go into them can incorporate millions of parameters connected in ways that are impossible to map. Not even the finest AI engineers can explain what exactly goes on in the mind of those algorithms as they produce specific outputs. It makes transparency all but impossible.
And a need for accountability in decision-making takes on epic philosophical tones when it comes to AI. Can an entity grown out of mathematics and data alone ever be said to be truly accountable? In what ways could a system found to be systematically creating errors be sanctioned? Could those errors ever be said to be on purpose?
Until such weighty matters are figured out, both law and market are likely to steer clear of depending too heavily on AI for accounting.
The Impact of Artificial Intelligence in Accounting Education
Getting a grip on the good and the bad in accounting AI will require a strong educational background.
Earning a CPA is already a process that involves an extensive amount of study in areas like:
- Auditing
- Tax accounting
- Managerial accounting
- Federal income tax requirements
- Ethics
Although it’s not on the exam yet, for all practical purposes artificial intelligence should be on that list too.
No one is expecting even top-flight CPAs to be experts who can code their own machine learning algorithms. But they will be expected to understand the basics of how those algorithms handle data, what the strengths and weaknesses of AI are, and how to use it as a practical tool in various accounting scenarios.
Every accounting degree program already includes coursework on computers and software used in the field. Many of those classes are probably already covering AI developments that are trickling out. In other cases, universities are beginning to offer specific elective AI in finance courses.
A few schools are going even further, offering concentrations in Artificial Intelligence in Accounting and Assurance with degrees like the Master of Accounting in Financial Accounting or an MBA in Professional Accounting. AI also features prominently in degrees like a Master of Science in Accounting Analytics.
These dive right into specific uses of AI in auditing analytics, assurance, and even new areas like continuous auditing.
Graduates from these programs will be the decision makers that determine how automation and artificial intelligence in accounting will progress. AI in finance and accounting will change the industry. But it is still up to humans exactly how that comes about.
In any case, the future of accounting will be unavoidably entwined with artificial intelligence. Accountants who study for that future today will be the ones who do best when it arrives.
