By Christopher Niesche

Machine learning and artificial intelligence (AI) have hit a tipping point where they can start to take over significant parts of corporate tax reporting and returns, says PwC partner David Lamb CA.
But the change isn’t happening as quickly as it should, says Auckland-based Lamb, who will be speaking on “Where to now with technology and tax?” at the CA ANZ Tax Conference in Auckland on 21-22 November.

“There are lots of priorities and it’s all about where they sit. Sometimes, with all due respect, tax is not the top priority of corporates and the people in organisations that make the decisions, like the CFO,” he says.

Machine learning will take over one of the key and most labour-intensive processes of preparing a tax return – recategorising entries in financial statements to entries suitable for a corporate tax filing.

“Machine-learning data tools can now identify and classify as a human would, but more quickly and more accurately,” says Lamb. That vastly improves the efficiency of the process.

“Machine-learning data tools can now identify and classify as a human would, but more quickly and more accurately.”

Machine learning leads to better business insights

Picture: David Lamb CA.

The transition to machine learning will be an important step for accountants in corporate finance, as they will be able to shift resources from transactions processing to providing business insights and partnering with the business.

Artificial intelligence and analytics tools will be able to perform checks across thousands of transactions on a scale that wasn’t previously possible.

“Where targeted analytics tools can really help is quickly doing some checks and balances on those transactions, because often you accept the accounting numbers at face value and say that forms the basis of our tax,” Lamb says.

“But now revenue authorities are saying ‘prove to us that the underlying transactions are sound, we want to look at that in much more detail’. And the way we can do that is using these interrogative data and analytics tools.”

The process can also provide insights for the broader business that go beyond tax compliance, because tax is one of the few places that provides a good lens over all of a corporate’s numbers.

Automating several processes brings bigger savings

Lamb says the technology has reached a tipping point where it is sufficiently advanced for corporates to start deploying it in their tax function, but a question remains over how many will in the near term.

Many corporates are still relying on manual tax returns, involving teams of finance professionals and all the inherent human errors.
“The question is where do you start and how do you assess the cost and the benefit of using the technology? Because, like all things, there is a cost to using technology and there are benefits,” he says.

Eliminating a single manual process on its own might not be worth the investment, but the savings soon multiply when several processes can be automated, says Lamb.

AI and machine learning will change the corporate finance function

The implementation of the new technologies will not only transform the corporate finance function, but also the jobs of those driving the change.

“What’s required of the tax professionals of tomorrow is to really have an ability to bring together technology, plus the best in processes, the best in people and in culture,” Lamb says. “It’s about really having the leadership skills to manage those elements.”

“What’s required of the tax professionals of tomorrow is to really have an ability to bring together technology, plus the best in processes, the best in people and in culture.”

Despite the potential of the new technologies, Lamb offers a caveat: “What's also important is for tax professionals and corporates to not just automatically trust the technology.

“How do you know it’s producing the right outputs? What steps do you put in place if the technology breaks down or things change? And what governance do you put around it, and what controls and processes?” he asks.