On a continuous basis, articles in news media outlets warn of job obsolescence, thanks to the ‘threat’ of artificial intelligence (AI) and automation. Some go on to highlight the occupations most at risk – and soon-to-be-made-redundant accountants are almost always a feature.
This fatalistic view of human accountants’ futures could very well be overstated. In the US, for example, the demand for accountants is set to grow by 16 percent through 2020. A 16 percent job growth rate through 2020 is also projected for auditors. In Singapore, accounting was amongst the sectors expected to see the strongest job growth in 2019, with 2,000 new jobs set to be created by 2020.
Why AI Accounting Can’t Replace Humans
Going against the intuition that automated, AI-powered accounting software may make humans dispensable, there are three reasons why we still need human accountants.
1) Human accountants are the translators of financial data and jargon
Accounting is a finance ‘dialect’ that takes years to learn. The more competent ones are Certified Public Accountants (CPAs) or Association of Chartered Certified Accountants (ACCA). The certification processes needed to acquire these titles are known for their robust syllabus, difficult exams and low success rates. As a business owner, when you have qualified accountants on your payroll, you can be assured that they can translate raw financial data into sound insights that you can understand and use for business decision-making.
2) Human accountants deliver deep analysis that helps business planning
As alluded to above, raw data is not useful for decision-making and you will still need a specialist’s knowledge to interpret financial information for business planning. At this point, no form of AI can explain how quantitative data relates to a company’s unique structure and operations and then provide qualitative suggestions that cater to evolving business circumstances, needs and goals.
For example, a small business owner who runs an ice cream parlour knows that his milk would typically last for seven days before starting to turn bad. The milk supplier has a delivery turnaround of three business days and has just introduced a bulk purchase discount of 15 percent. The volume of visiting customers is twice as high on weekends, as compared to weekdays, but that again changes on public holidays or during school holiday periods.
The above represents raw data that the ice cream parlour owner has gathered from running the business. An accountant would then add value by applying techniques like calculating Economic Order Quantities (EOQ), implementing cashflow management and identifying the right days to place optimal restocking orders that match inventory supply to customer demand.
3) Accounting is subjective in nature
While one might argue that enterprise resource planning (ERP) software with built-in accounting and inventory management modules can accomplish the above, we need to acknowledge that the software is merely a technology tool and not a human replacement.
There is usually more than one way to interpret an accounting principle, which means that you can choose the way you want to account for certain things. With new and updated guidelines constantly being released by boards that determine and oversee financial accounting standards, there are also infinite ways for reporting and analysing financial data, before arriving at a decision.
Consider this simple example – a business owner taking up an annual trade association membership for $10,000 a year is presented with the options of paying the entire fee upfront or making monthly instalments of $850. The business owner would make a choice depending on whether the business has a tight cashflow. If the business has good cashflow, it would be advantageous to pay the lump sum and save in total costs. If cashflow is tight, it might be preferable not to add excessive financial burden in the immediate term and pick the instalment payment method instead.
Even if the software can help to decide based on pre-programmed rules, a human must still configure the machine based on his or her preferences. AI may be able to predict future cashflow based on historical trends and data, but it is unable to replace the business experience and judgement that a business owner or an accountant can provide.
Ultimately, technology can help to unravel viable options quicker, but there is no substitute for the ‘last-mile analysis’ and final decision made by humans. And at this stage, it is more art than science.
How Then Will AI Transform Accounting?
In basic terms, AI is mathematical algorithms designed, built into software and trained using data. The technology can complete repetitive and mundane tasks at a fraction of the time it takes humans and with greater accuracy. The emergence of machine learning now allows AI platforms to observe, analyse and self-learn data and processes to improve the software’s performance and accuracy over time.
While leading accounting software providers like SAP have incorporated AI and machine learning into their solutions to handle basic accounting tasks, such as bank reconciliations, invoice categorisation, risk assessment, and audit processes like expense submissions and invoice payments, the narrative of automation killing jobs is a misguided view. This view is in fact, more than half a century old.
The rise of ATMs in the 1970s, for example, did not cause the anticipated massive unemployment for bank tellers. Rather, it fulfilled the ‘automation paradox’ and made bank branches cheaper to operate, which then led to more branch offices being opened in underserved territories and a higher demand for bank tellers.
The ATM also transformed bank tellers’ roles and made them more customer service orientated. Today’s tellers don’t just accept cash deposits and fill out money orders, they have become the frontline providers of advice and a personalised experience that is crucial for customer retention. These days, they are further empowered by automated and AI-powered customer relationship management (CRM) technology to determine what conversation might be the best one to have with an individual.
Similarly, rather than replace the human accountant, AI will transform the duties a human accountant performs. And we could therefore make a case for AI to stand for ‘augmented intelligence’ instead.
AI as ‘Augmented Intelligence’ in Accounting
With AI and machine learning handling many of the mundane, repetitive tasks, accountants will have more time to focus on other aspects of the job. Take a company’s accounts payable department as an example. An accountant in this department works eight-hour shifts and must respond to customer and vendor inquiries that rain down on his or her to-do list every day.
After struggling to identify the dilemma posed in an inquiry, the accountant must find the related reference documents across an extensive but heterogenous filing system. Too much time is spent searching for documents and becoming frustrated as more and more requests enter the accountant’s inbox.
With the help of AI, robot process automation (RPA), and chatbots, the company can help the accountant by enriching its processes with AI and increasing the level of automation.
Machine learning-based document processing services delivered by SAP Intelligent Technologies, for example, comprise a powerful technology to detect and highlight any type of named entity in unstructured text and sort it into predefined categories. By doing so, this service helps to increase overall document-processing productivity, increase efficiency, and reduce errors.
In the context of our example, the AI algorithms automatically extract unstructured information, as well as learn from historical datasets and the interactions of the accountant with the system, thereby improving the matching rates of inquiries to relevant documents over time. In the end, the solution can analyse incoming inquiries automatically to assist the accountant or even directly answer simple, straightforward inquiries on his or her behalf.
In instances like this, the human accountant will still be relied upon to provide emotional intelligence, excellent communication and, above all, honesty and integrity when imparting advice and making decisions. Rather than expound hours completing menial tasks, accountants will also be freed up to explore the more creative realms of their roles and extend their skills in problem solving, advising, strategy, and leading.
Ultimately, AI is not displacing accountants. Rather, it is augmenting their technical proficiencies and critical thinking skills to make more insightful decisions, while increasing the efficiency of producing accurate financial statements and reports. Accountants of the future will just need to adopt this technology – just as they had to adopt the smartphone, the computer and the internet.