The Big Four do not intend to replace human accountants with artificial intelligence anytime soon — and even if they wanted to, firm leaders concede that the technology isn’t yet at the level where they could contemplate doing so. This is despite each of the four largest firms making billions of dollars of investments in the sector over the past year.
Instead, the focus at each of the firms seems to be more on augmentation versus full automation. Cliff Justice, KPMG U.S. enterprise innovation leader, said right now the firm is thinking of AI in terms of increasing the efficiency and productivity of human accountants, enabling them to concentrate on broad strategic thinking while computers take care of routine tasks like data entry.
This is because accounting is more than these routine tasks; despite advancement in AI, Justice believes the human element will remain essential in the profession. “It won’t replace the human at the human-to-human interaction level. It doesn’t ideate, it doesn’t create innovations by itself. That may come in the future — I’m not one to say technology can’t do something because you can always be proven wrong — but at this stage it is very much augmentation for our people and support for our people,” he said.
Paul Goodhew, global assurance innovation and emerging technology leader at EY, raised a similar point, saying it’s not so much about replacing the human but, rather, enhancing and amplifying the capacities they already have. For instance, AIs can process data much faster than people, and so can “interrogate the data more robustly” in a way that enables the accountant to ask better questions and gain better insights. But, he said, the human will remain central because it is only the human who can make the appropriate judgment calls based on the data.
“We recognize it will be essential for the human to always be there to provide the deep judgment and provide that experience and skepticism. Ultimately the AI might help them get answers, but the human judgment is essential,” he said, adding that he sees humans and AI complementing each other.
Chris Griffin, regional managing partner for audit and assurance at Deloitte, raised this same issue, noting that while AI is a powerful tool for automation, human judgment and skepticism will remain paramount.
“Our talent is our most important asset, and this doesn’t change that. It may change what some of our talent does on a day to day basis, it may automate certain tasks, but it won’t be end-to-end ‘fire and go’ kinds of results because you still need that judgment, that skepticism, to deal with things like bias and reliability, those kinds of issues core to what we do,” he said.
Joe Atkinson, chief products and technology officer with PwC, said it’s not necessarily an either/or formulation. Like the others, he said AI is not a replacement for accountants but a capacity builder, and emphasized that PwC’s AI strategy is “a human-led strategy” that involves equipping people with the best technology and the education to use it well. He recommended against thinking in terms of all-human or all-machine, noting that there is no AI without data, which is primarily generated by human activity.
“I think there’s a role for both. I’m not a purist on one side or the other. But I think the power of augmentation, ultimately, will unlock the power of automation by itself. We’ve been on the automation journey [for many years] and we’ve benefitted enormously,” he said.
Not ready for prime time
Even if they wanted to replace human accountants with AIs, all four said that despite impressive advancements over the past year, the technology still isn’t there yet to do so. At the moment, at least, it is just not technically feasible.
“All that technology is not developed. All those models haven’t been trained. Theoretically, from a technical point of view or from the point of view of what we are required to do as an organization when it comes to delivering services, it wouldn’t be applicable,” said EY’s Goodhew.
Will Bible, Deloitte’s digital transformation and innovation leader, noted that in certain cases — such as audit — the notion of having an AI perform the entire engagement would seem to defeat the entire purpose. When people hire auditors, he said, one of the key elements of the process is the assurance and responsibility for the professional’s opinion.
“Right now, a machine cannot take responsibility for an audit opinion. It requires a person. From the conceptual level, an audit product, at its most fundamental nature, must have a person involved, who is an experienced qualified professional,” he said, but added that there is certainly a large role for AI in the audit that does not involve issuing an opinion.
But even if the technology were there to completely replace a human on an engagement, PwC’s Atkinson wouldn’t want it to.
“I would say the technology isn’t there yet. And I would also, and maybe more importantly say, I don’t love a vision of allowing a computer to do it end to end. This is where you get into the trust, integrity and responsible use of AI. The question to me is about what clients ask for, what they count on us for: the application of professional judgment. AI tools are really good at pulling out information and making predictive choices but they can’t replace human judgment,” he said.
Credit: Source link