5 ways accountants have mastered AI

What accounting tasks are likely to be affected by AI?

The new way of thinking about AI is to see it doing time-consuming tasks, freeing up space for accountants to do the serious thinking and to exercise professional judgement on more complex matters. Despite some people’s fears, AI’s advocates say it can be a job-creator, not a job-killer.

For years, there have been fears that Artificial Intelligence (AI) – smart machines that work and react like humans while having self-learning capabilities – will redefine the role of accountants.

Now innovative firms are investing in AI so they can be at the forefront of cognitive technologies. What does the evolution from automation and data-analytics software to AI mean for accountants? It isn’t about to deliver a robot takeover of the economy but it will bring change. 

Here are five accounting tasks that are likely to be affected.

1. Auditing

KPMG accountants and their clients will benefit from the wisdom of a rather unusual partner – Watson the supercomputer – now that the Australian arm of the firm has signed a deal with IBM to use the famous cognitive system.

Integrating machine learning and other AI technologies, Watson can pore over thousands of pages of contracts or documents and quickly summarise them into useful criteria. It can check and crossfoot debit and credit entries and allow analysis of larger populations of data than traditional sampling, freeing up valuable auditor time to focus on areas requiring significant judgement and expertise.

Ken Reid, national managing partner, brand and innovation at KPMG Australia, says specialised software already automates a lot of structured data relating to auditing tasks.

“The difference in the cognitive world is that you can then ingest unstructured data from multiple sources and look for trends and correlations,” he says. 

Deloitte, among other firms, is following suit and has joined forces with machine-learning platform Kira Systems to transform contract and document reviews.

At KPMG, piloting and prototyping of services drawing on Watson’s skills is well advanced and the rollout of some services has started. However, Reid says some of the most complex AI innovations across audit and other services may take anywhere from months to years to materialise.

“We certainly don’t see it as robots taking over. We see it as using technology in different ways to augment the skills of our people,” Ken Reid, KPMG

“It’s not a quick process,” he says. “(In some cases) we’re trying to teach Watson to do something that a professional partner with, say, 30 to 40 years of experience can do.”

Although some fear job losses, Reid believes Watson, and AI more generally, will cut out some of the mundane process work as talented accountants work in tandem with supercomputers on more innovative or sophisticated tasks. Contemporary auditors who understand and can enhance cognitive systems will be in demand.

“We certainly don’t see it as robots taking over. We see it as using technology in different ways to augment the skills of our people.”

2. Risk management

With its focus on numbers and data, the accounting sector – and areas such as risk management, in particular – is a prime candidate for AI’s intelligent learning systems.

“Any tasks which are mostly repetitive are open to change,” says Mohit Sharma, managing director of Mindfields, an outsourcing and automation advisory firm.

Just as Uber has revolutionised the concept of sharing cars for transport, he expects more and more technology firms to share their knowledge-based assets with firms of all sizes, allowing them to easily access the best technology on a pay-per-use basis or at a relatively low cost.

Professional Development: Want to learn more about how AI and analytics are shaping the future of business? Prof Michael Davern CPA, chair of Accounting and BI Systems at Melbourne University, will speak at CPA Congress 2018 on the topic. Find out more.

The domain of risk management is well suited to cognitive computing capabilities because fraudulent cases can be missed amid mountains of structured data or ambiguous events.

With AI, natural language processing uses algorithms to analyse text and uncover discrepancies with tax payments, for example, while also having the capacity to self-learn and avoid false-positives if previous suspicions of fraud were proven to be incorrect.

3. Vendor reconciliation

Cloud-based providers such as German accounting software startup SMACC is using AI to take the automation of tasks, such as invoice processing, to a new level. Customer receipts can be turned into a machine-readable format, encrypted and then allocated to an account. The platform can also self-learn while tracking invoices and sales and costs data.

Annie Flannagan, founder of Better Business Basics, which has automated the bookkeeping side of her business, says there is no doubt the accounting sector is “ripe for disruption”.

However, she draws a distinction between automation through popular platforms such as Receipt Bank and true AI.

“Often what we look at and label as AI is actually technology efficiency and automation,” Flannagan says.

AI is about knowledge, not just data, and the real benefit for accountants’ clients will stem from taking multiple software systems and letting them talk to each other. “What is missing is the integration piece,” says Flannagan, who adds that a combination of smart AI systems and intelligent people and processes has the potential to transform accounting.

Although AI is likely to create rather than kill off jobs, Flannagan says accounting firm leaders must foster different jobs for people who have the skills and flexibility to adjust to ever-changing technology forces.

“You’re not necessarily putting people out of jobs, but you are putting them on notice. If you’re walking into a finance role that you expect to be static, well, those roles are not going to exist in the future.”

4. Regulatory compliance and reporting

Accounting firms seem set to follow the compliance lead of the big banks, which are using AI technology to navigate an ever-changing regulatory landscape. 

AI lets wealth managers, for instance, handle data about investment portfolios and transactions while giving compliance departments access to more information that helps them understand the impact of new rules.

Within accounting firms, intelligent machines will be able to monitor compliance with regulations as well as organisational policies. There should also be an ability to stay on top of frequently changing tax rules and identify suspicious transactions more quickly in areas such as intellectual property theft and insurance claims.

“The real benefit of such technology advancement will be the reduction of costs for the consumer,” Mohit Sharma, Mindfields

As machine-learning options grow, clients will have access to a massive amount of data-driven knowledge at increasingly affordable prices. 

“The real benefit of such technology advancement will be the reduction of costs for the consumer,” Sharma says.

He adds that the rise of AI will complement the evolution of accountants from number crunchers to value-added advisors.

5. Trend analytics

Perhaps the most exciting aspect of cognitive technologies for accountants will be the ability to reveal insights that can inform operational and strategic decisions for clients.

AI and machine learning will let them take large and complex data sets and make high-quality predictions from that data, potentially giving clients an edge in their field of business.

Flannagan says such information, along with real-time benchmarking services that help businesses compare how they are performing against their peers and similar industries, could be a game-changer for smaller businesses that previously have not had access to such technology. “That’s when you start to get much smarter, rather than comparing yourself to yourself,” she says.

Meanwhile, as KPMG accountants get acquainted with their new colleague Watson, Reid says the broader focus is on how data analysis using supercomputers can transform the way critical business decisions are made. He notes that Watson even has the capacity to assess the sentiment and tone of communications such as voicemails.

“It’s an exciting evolution,” Reid says.

Read next: Here's why robots won't take over the economy anytime soon


Like what you're reading? Enter your email to receive the fortnightly INTHEBLACK e-newsletter.

December 2018
December 2018

Read the December issue

Each month we select the must-reads from the current issue of INTHEBLACK. Read more now.

CONTENTS