There’s some understandable concern these days about the possibility that increased use of artificial intelligence (AI) will lead to job losses among accounting and audit professionals.
But Mike Baccala, PwC’s U.S. assurance innovation leader, predicts that while accounting and auditing jobs may change as AI use becomes more prevalent, those jobs won’t go away. Baccala is co-author of a recent PwC report that predicts the effects that AI will have on the business environment.
He said PwC already is working with AI in a few capacities in client engagements. The firm is using an AI platform to help nonaudit clients extract data from their lease agreements as they implement the new FASB lease accounting standard. Without AI, this extraction would take eight to 10 hours to perform for each lease contract, and some clients have thousands of lease contracts. With AI, that time is down to three or four hours per contract, and it’s continuing to decline.
In addition, PwC financial statement auditors are using AI to draw data out of client bank statements to help with substantive testing that’s required for auditing of cash. In both cases, Baccala said, the AI hasn’t resulted in people being taken off the projects; instead, the people are performing different duties.
The AI is pushing people toward performing work that is more interesting, Baccala said. For a long time, one of the more tedious tasks for accountants and auditors has been the process of taking data and organizing them before they are analyzed and audited. Those data could take the shape of a bunch of receipts and invoices in a shoebox, or the form of various ledgers and spreadsheets.
AI can replace humans at the dull task of extracting, organizing, and structuring the data. But those same accountants and auditors working with AI perform different tasks. First, they teach the AI what data to look for and how to organize them. Then they investigate anomalies. And because the AI is working with all data rather than just a sampling (in the case of audits), there may be more anomalies to investigate.
Meanwhile, the AI-enabled auditing of all data rather than just samples is providing more value to clients.
“We haven’t actually removed anybody from the practice [because of AI],” Baccala said. “We’ve freed them to do other things, and we’re helping more clients with the process.”
Baccala acknowledged, though, that a significant amount of “reskilling” needs to take place to help accountants and auditors work effectively with AI. He said traditional CPA skills such as skepticism, judgment, analysis, and understanding technical accounting will become even more important because those are qualities that AI is unlikely to ever duplicate and, in a data-governed world, they are vital.
New skills that help accountants and auditors succeed with AI include:
- Fundamental data skills. “If you look at every client and the variability in their own systems, processes, data, policies that affect that data and how the accounting gets done, it’s a massively complicated web of different variables that can change from client to client,” Baccala said. The most successful accountants and auditors will possess core data skills that allow them to succeed as they serve clients with different systems and policies. These include data strategy and data processing skills, as well as proficiency in statistics, probability, and deductive reasoning.
- Storytelling ability. PwC is focused on training its people in the art of using data to effectively convey meaning and a message through storytelling. The AI can provide huge amounts of data, but that’s useful only if accountants and auditors understand how to translate the data in a way that makes sense to their audience.
- Ability to automate. Accountants and auditors are likely to encounter processes used by clients that could be improved through automation. Whether those are production processes or finance processes, the ability to implement automation to improve efficiency and reduce costs will be a differentiator for CPAs.
Baccala cited studies pitting chessmasters against machines to explain that people and machines working together have been shown to be much more effective than either people or machines working on their own. He sees technology, people, and processes as a three-legged stool essential for success in any business. If you cut one leg short, the stool won’t support much weight.
So while leaders in business and the accounting profession embrace technology, they should be careful not to neglect their workforces and their processes, Baccala said.
“You cannot leave the people behind,” he said. “Any organization that’s trying to do this isn’t going to get very far with AI if they’re not upskilling their people and using people as part of the process to move forward.”
—Ken Tysiac (Kenneth.Tysiac@aicpa-cima.com) is a JofA editorial director.