“We are an IT services company. Hiring highly qualified, talented team members, globally and at scale to support our own customer demands is our job, that’s our business,” said TELUS CFO Vanessa Kanu, during Fortune’s Emerging CFO discussion, a virtual conversation on Wednesday that focused on ways to address the talent gap with advanced technologies. The Canadian tech firm was able to cut down the time from application to onboarding by “well over half,” according to Kanu, by leveraging the metaverse for talent acquisition.

The proliferation of A.I. and automation-driven tools has caught the attention of organizations across the globe. Adoption of A.I. tools has more than doubled since 2017, according to research by McKinsey, but it has hit a plateau the past few years.

“Finance is in danger of becoming real laggards in the area of A.I. and automation, and even traditional analytics,” said Tom Davenport, president’s distinguished professor at Babson College and author of All-in on A.I.: How Smart Companies Win Big With Artificial Intelligence.

Still, he sees a lot of opportunities for the finance industry to incorporate A.I., nothing that robotic process automation is not seen as the most exciting A.I. tool, but is quite useful, as well as simple machine-learning predictive analytics.

Generative A.I., Davenport explained, could be used at financial firms like hedge funds to create spreadsheet formulas. That could ultimately save a lot of time on labor-intensive work and ultimately require fewer lower level analysts at those organizations. As is always the case with A.I., this raises questions about the disruption of elimination of jobs, though conversely, it could give employees an opportunity to take on more work that requires strategic thinking.

“If you freed up people from just recording-related activities and used them to start doing analysis, prediction, and trying to understand what are the factors that drive our performance and doing a better job of predicting what future financial performance is going to be, that would be an obvious substitution,” said Davenport.

Not all financial tasks can be automated, said Davenport. At the end of the day, someone needs to be held accountable for financial performance, and that includes the CFO and external auditors. An A.I. system cannot be sent to jail if things go wrong.

“A.I. and the insights are only as good as the underlying data,” said Katie Rooney, CFO at Alight Solutions, an Illinois-based IT and consulting company. “Our first focus has really been streamlining that process.”

All of Alight Solutions’ systems—from HR to finance—now on Workday, which has helped level set the teams. The next phase, said Rooney, is leveraging some of that data to actually drive decision making, and that’s where the human element comes into play. “A.I. is not a replacement, it should be a tool,” said Rooney. “It helps us work differently, provide insights differently, and free up capacity.”

The increased use of automation even raises the question of if Alight Solutions itself could be eliminated. After all, it uses data-driven tech to help customers navigate health, wealth, and payroll. Couldn’t a machine simply do that?

“It’s a fascinating question I am starting to hear from investors: ‘Well, will you still be around?’” said Rooney. But she goes back to the underlying factors of what Alight Solutions does: they help 36 million people today and can help customers navigate what their health plan is and match what doctor can help best address their needs.

“I think A.I. will continue to help and transform the way we can look at data,” said Rooney. “But there is still the human element in how we help people make a different decision.”

Kanu agrees with Rooney that with the implementation of A.I., it must start with data. TELUS operates in 30 different countries with 75,000 employees and with that level of complexity, different departments across the organization can quickly start speaking different languages.

A.I.-driven automation can support TELUS’ employees, but those decisions need to be data-driven, said Kanu. She points to the success of an internal resource chatbot tool known as “Sara,” which today helps address half of employee inquiries, and thus eliminates some repetitive tasks that were previously handled by the HR team.

Kanu said that ultimately, anything that requires human judgment—ethics, decision making, relationship building, or addressing key stakeholders—won’t ever be eliminated by automation.

Read the full report here