Is Artificial Intelligence (AI) the Future of Accounting?

Digital rendering of a Human Brain Covered with Networks, representing artificial intelligence.

Dong Wenjie / Getty Images

Smacc, a German-based software firm, uses artificial intelligence to help freelancers, small companies, and medium-sized enterprises automate their accounting systems and financial reporting. They received $3.5M in Series A financing from a variety of high-profile venture capitalists and angel investors, and the founders developed the concept after experiencing difficulties with accounting in the initial stages of their own startup company.

Smacc clients transmit their receipts, which are then converted into machine-readable form. The receipts are allocated to the proper account after encryption. Over time, the system teaches itself to improve its functions: sales, expenses, invoice management, and liquidity profiles.

Self-Learning and Improvement

The software uses more than 60 data points to review receipts and invoices. It checks whether the math is accurate and verifies whether the issuer is correct with details like Value Added Tax (VAT) identification numbers. When the software has learned how to handle each supplier, tasks are subsequently handled automatically. Its artificial intelligence allows it to self-learn and constantly improve its ability to sort and allocate information.

Customers can check their billing and expense data in real-time online, and no longer have to input data or wait around until the end of the month to see where their finances stand. Several companies, such as QuickBooks, offer cloud-based accounting software, but Smacc is among the first to leverage artificial intelligence to improve the software’s ability to automate tasks.

The Rise of AI

The world of accounting is just the latest in a series of industries being affected by the rapid increase in the use of artificial intelligence. Bill Gates even referred to the rise of artificial intelligence as computer science's "Holy Grail." After many failed efforts in the past, the accuracy and speed of today's artificial intelligence are much improved.

You cannot go a day without someone on your Facebook feed sharing an article about artificial intelligence and how it will take your job in the next few years, but these concerns are not new. The same fears were at the forefront of people's minds as factories spread throughout Britain 200 years ago. 

Robots are already used throughout our homes, workplaces and entertainment centers, and over the next 10 years, Forrester Research estimates that AI will take over up to 16% of jobs in the United States. Google believes that robots will achieve human intelligence levels by 2029, and Gartner estimates that 33% of all occupations will be performed by smart robots by 2025. FOW predicts five areas will feel the most impact: healthcare, manufacturing, transportation, customer service, and finance.

Reality Check

With all that said, accountants more than likely do not have to worry about artificial intelligence for a long time. Smacc is developing interesting AI applications to help further automate and streamline bookkeeping tasks, and cloud-based accounting software packages such as QuickBooks say they are already 75% automated. That said, Professional accountants do much more than keep track of receipts and provide basic reports. They act as consultants who advise on tax planning, discuss operations, review client goals, and more. The rapid pace of change in client industries and the expansion of complicated regulations means that human controller services will be necessary to ensure that compliance requirements are met and financial controls are sound.

Multiple Countries

This is especially true for companies that operate in multiple countries. It is difficult enough to deal with taxes in your home country, but making sense of the tax code and business regulations in a number of foreign countries is daunting. Are AI robots ready to deal with the tangled web of regulations associated with the European Union or the compliance requirements of the Organization for Economic Cooperation and Development (OECD)? No artificial intelligence algorithms that can sort out these complex interactions currently exists. 

Context Challenges

Machine learning can be trained to handle an amazing variety of tasks if you give it a wide enough variety of examples to draw from. Data scientists are not exactly sure how this happens. The math is so complex, it's difficult to re-engineer it to see how the system learns, which makes diagnosing problems difficult.

AI can do amazing things, but it's not so good at the many things humans do naturally. We make a lot of decisions based on context. Professional controller services understand the rules and regulations their clients must adhere to, and they're able to present options and recommendations in a manner the client can understand.

Present-day machine learning systems don't handle this type of context well. Futurists have proclaimed the benefits of AI for decades now, describing amazing worlds where robots make your everyday life one of ease and relaxation. That future may be here faster than you think, but for now, outsourced accounting services have an advantage the most advanced algorithms cannot duplicate—the human touch.