Accounting | Artificial Intelligence

Accountants can only gain from AI, say experts

Will robots replace CPAs? The short answer is no, if we learn to embrace artificial intelligence and the opportunities it presents. Here’s how you can gear up for the future.

A Facebook IconFacebook A Twitter IconTwitter A Linkedin IconLinkedin An Email IconEmail

Silhouette glasses in front of laptop computer screen with programming codedCPAs will need to understand coding languages such as SQL and Python in order to manipulate and transform data into insightful analytics (Shutterstock/Zapp2Photo)

Chatbots, self-driving cars, personal recommendations from Netflix: these days, it’s impossible to ignore the amazing advances that are taking place in artificial intelligence (AI). But amazement at those advances is also coming with a lot of dread about the consequences for our job market. As Michael Wong, CPA, principal of research, guidance and support at CPA Canada, explains, “Now that AI is really taking off, everyone is talking about it, everyone is raving about it and everyone is scared of it.”

But should accountants be concerned about their jobs?   

According to Wong, who authored a CPA Canada paper, Big data and artificial intelligence – the future of accounting and finance, there is no doubt that AI will have an impact on the profession. But it won’t replace CPAs if they embrace it and make the most of the opportunities it presents. “Basically, AI is going to displace many tasks but it’s not necessarily going to replace jobs,” he says.

Jeff Lui, CPA, a director in Deloitte’s AI practice, has a similar take. 

“Organizations often are very extreme in their thinking, imagining that machines will run entire businesses,” he says. “But in reality, at least in the near future, they will simply make our lives easier and more efficient.”

Here are four ways AI will affect CPAs’ work, and what they need to do now to prepare.

1. A PRODUCTIVITY BOOST 

With the deployment of AI-powered tools, many routine tasks will be eliminated. It won’t be necessary to manually categorize expenses, because software with built-in AI will be able to learn from the way past transactions were categorized and automatically categorize future ones—even if they are new types of expenses. And drone technology powered by AI can help with inventory counts. That makes it unnecessary for companies to send their auditors in person to do a count in locations that could potentially be hazardous—construction sites, for example. 

Still, the use of drones or other productivity-enhancing tools won’t take away the need for auditors; they will simply be freed up to spend time on more value-added tasks. What’s more, professional judgment will still be needed. 

“A drone can count the inventory and say this is what I saw, but someone still has to verify the count,” says Wong. 

Lui agrees. “AI machines are not perfect,” he says. “They still rely on a large amount of data to ensure high degrees of accuracy, and in that spirit, humans should always be in the loop to maintain a level of certainty. That is where AI still lags behind human intuition—right now, I can show a photo of one dinosaur to a small child, and he or she will know what a dinosaur is. Conversely, a machine needs thousands, if not tens of thousands, of photos of dinosaurs to ‘learn’ what a dinosaur typically looks like.”

2. BETTER, FASTER AND CHEAPER PREDICTIONS 

With the growth in data from nontraditional sources, combined with advances in AI, it’s becoming possible to move into advanced analytics.

“With AI, you can analyze far more data, from more sources, to drive better predictions, than we ever could access using traditional methods,” says Wong. “Moreover, AI can find correlations hidden within the vast data that we might not even be aware of. For example, one company was able to determine, with the help of AI, that one of their yogurt flavours was less popular than others when the weather was hot. As a result, they were able to reduce production of that flavour during the hotter months and reduced spoilage.”

In this new AI-driven world, CPAs will be expected to evolve beyond their traditional forecasting methods to leverage the insight that AI comes up with. As Wong puts it, “CPAs will need to apply their judgement. You still need someone to act as a check and say, ‘AI told me this, but is it really applicable and does it make sense?’” 

3. CHANGES TO THE AUDITING PROCESS

Over the next five to 10 years, the term “audit” could well change, says Lui. As AI becomes more prevalent in organizations, accounting firms will move way from financial statement auditing as a major revenue stream and start to look at auditing algorithms and auditing data sets. “That will be a big source of revenue generation,” he says.

Because of the power of AI and machine learning, auditors will also be able to test an organization’s data in its entirety. “Say I have 500 documents from a client,” says Lui. “Instead of taking a sample, as most audits do, it will be possible to scan through all 500 documents word by word and get down to the nitty-gritty very accurately. We won’t ever be 100 per cent accurate that there is no fraud going on, but we can be certain that 100 per cent of the documentation and financial tests have been performed. However, there are still ways that humans can circumvent the system undetected by AI algorithms.”

Lui has also heard of firms doing real-time audits. “Instead of conducting an annual financial statement audit, they have a continual monitoring process. So you get real time updates, rather than getting a surprise at the end of the year.”

4. A NEED TO LEARN NEW SKILLS

Even though AI tools are built primarily by engineers, accountants are definitely part of the process. As Lui explains, an engineer can only take it so far. 

“You need to contextually understand what’s going in the algorithm,” he says. “And if you’re building an accounting machine learning algorithm, you will need an accountant there to understand the intricacies of building that system, because an engineer doesn’t have the context. The CPA knows what to look for, the fraudulent items that should be red flagged. It’s definitely a team effort,” he says.  

Wong adds that CPAs will need to learn how to access databases, because that is where the data they will be using to drive their analysis resides. They will also need to understand coding languages such as SQL and Python in order to manipulate and transform that data into insightful analytics. “I believe coding will be an important and fundamental skill for everyone,” he says. “It’s almost like a third language we need to learn—the language of machines.” 

For those who make the investment in acquiring that knowledge, the payoff will be huge. “Today, knowing, understanding and using AI can provide a competitive advantage,” says Wong. “But soon, that will all be just business as usual. It’s not too late to start learning now.”  

DEEPEN YOUR AI KNOWLEDGE

Big data and artificial intelligence – the future of accounting and finance is the second in a series of publications by CPA Canada focusing on AI. The first publication, A CPA’s introduction to AI: From algorithms to deep learning, what you need to know serves as a primer on AI, explaining AI “buzzwords” and discusses the evolution of data, AI and computing power.