AI practices within accounting: the risks vs. the rewards

Listen to this article

As artificial intelligence comes into greater focus for business, and provides big opportunities for efficiency, some potential pitfalls are looming, according to local accounting professionals who count Long Island organizations among their clients.

Onerous, repetitive tasks in areas such as accounts payable and receivable could become infinitely easier with the right AI-based technology. But without safeguards like encryption and thought-out policies, businesses could instead be introducing themselves to risk rather than reward.

“Some clients block access to AI applications like ChatGPT, while others are establishing committees to oversee its responsible use,” said Jennifer Wood, a CPA and assurance line leader for The Bonadio Group.

“AI is a buzzword in public accounting, and firms including ours are exploring how best to utilize it,” Wood adds. “We’re enthusiastic but cautious, mindful of data security and client
confidentiality.”

The implications, for accountants, go way beyond taxes.

According to report last year by the Journal of Accountancy, “about half of all tasks U.S. workers perform could be completed faster by generative AI without losing quality.” It noted how “many accounting firms are open to more automation.”

Almost a quarter of client advisory services practices have deployed AI, according to a 2022 survey, the journal reported.

“We are arguably on the cusp of one of the greatest technological shifts of our time with AI… Chat-based services such as ChatGPT are incredibly intuitive and easily available and affordable for all businesses,” said Daniel Bridges, director of innovation at Mazars.

With these technological advancements aside, Bridges adds that it should still be approached with cautious optimism. “We should, however, recognize that this really is in its infancy and most businesses are still fathoming out how to get the most value from the potential,” Bridges said. “On the one hand, the technology can improve efficiency and accuracy, but there is also the potential for it to impact quality negatively, too.”

Professionals say accounting firms are now in the “exploratory” phase of understanding AI, and many of their clients haven’t even moved that far. Hard-and-fast guidelines by accounting trade groups have yet to take on the latest advances like Chat GPT, best practices in the use of AI-based applications, or create written guidelines.

But for many, it’s becoming more difficult to deny the power the technology brings to bear in accounting, auditing and related practices.

“Internally, AI is revolutionizing
efficiency, especially in tax and audit practices. For instance, optical character recognition technology allows for the rapid processing of client documents into tax forms.” said Al Borghese, a director at Cerini and Associates in Bohemia.

That also means for repetitive functions like accounts payable, software can scan and review documents, extract necessary data and compile it into applications that can speed up payments—with far fewer hours to get it done.

Rather than reducing the size of the workforce because of AI efficiencies,“it’s actually making the workforce, I think, rely less on those clerical kind of (functions) and they can start using their other skills—analytical type skills—that can drive a business more,” Borghese said.

According to Bridges, the integration of AI into workflows, like summarizing meeting notes, necessitates changes to our current practices: meetings must be recorded for transcription, and clarity in language and assignment of tasks must be enhanced for accurate capture by AI. To maximize AI’s benefits, we must adapt to more structured and precise communication, and learn to effectively direct AI with carefully formulated prompts.

“All of this requires investment in technology, training, and most importantly, time,” Bridges said. “Accountants must be given the time to experiment, test and refine these activities.”

And for a number of AI functions, human review is still a necessity. Consider recent notable cases whereby attorneys relied on AI to write legal briefs to submit to a court, only to learn the application supplied fictitious legal precedents in some cases.

For auditors, AI can examine data to flag potential areas of fraud—greatly expanding their field of vision. But the audit function will still require trained professionals that can go beyond the black and white to understand why numbers may not add up, experts say.

Industry professional groups The American Institute of CPAs and CPA Canada jointly published a report last year called “The Data-Driven Audit: How Automation and AI are Changing the Audit and the Role of the Auditor.” It’s conclusion: While the technology can make many aspects of work easier, it’s not a silver bullet.

“It is important to see automation, analytics and AI for what they are: enablers, the same as computers,” the report found. “They will not replace the auditor; rather, they will transform the audit and the auditor’s role.”

For now, while businesses should consider policies and implementation plans—common sense is an ally.

“(AI) is not going to give you the (perfect) answer, but it’s going to give you a darn good first draft and help you maybe consider things that you might not have initially considered,” Wood said. Having a human backstop in many AI uses, to keep an eye on the output and make sure it’s accurate and useful, is key.