Manage equity without headaches. Running a company is a complex skill. Managing your cap table shouldn’t make it harder. Growing companies use Pulley to simplify equity management from fundraise modeling to tender offers. Master running your business with help from Pulley.
What implications will a shrinking pool of jobs have on the finance function? What will work look like when so much is automated? Learn more about why you need to believe the AI hype in part 1 and part 2 of this article series.
The 2025 Gartner CFO & Finance Executive Conference had more than its share of chilling predictions about AI and how it’s displacing people. With so much being automated, what are the humans in finance still doing? What will the workforce look like, especially where new hires are concerned? And how is finance’s role within businesses going to change? Experts and finance leaders alike shared their predictions with CFO Brew.
Everyone’s a supervisor now: By now, everyone’s probably heard the standard line that “AI frees up staff to do higher-level work.” But for Brex’s accounting team, at least, the spiel is true. Instead of doing “grunt work,” staff are now reviewing the work that AI performs, Erik Zhou, Brex’s chief accounting officer, told CFO Brew.
Before AI, Zhou estimated, staff were manually coding 30% of the invoices that came in. Now that AI is able to reconcile as much as 95% of invoices with purchase orders, staff can focus on reviewing instead. “Now you’re doing a review of all the work from a tool to make sure it all makes sense,” Zhou said. “And, frankly, it is more efficient.”
Similarly, instead of cranking out flux analyses for individual vendors, employees have AI create them, then review them.
“And then they’re spending more time in the business investigating what’s causing the trends,” he said. “We also have a better financial picture across the whole team.”
Zhou likens AI to a junior coworker that employees need to supervise. Unlike, say, Excel, AI can hallucinate or drift, so staff need to pay close attention. They’re responsible for AI’s outputs, he said: “It’s [got] your name on it, and you have to make sure the result makes sense."
Are newbies SOL? (Or do they just need SQL?) That, of course, raises the question: How are entry-level workers going to get the experience they need to supervise AI if AI is doing that work for them—or if they’re not getting hired at all?
No one we spoke with or listened to at the Gartner conference had a clear answer. Deirdre Ryan, global finance transformation leader at EY, did tell CFO Brew that college accounting programs are starting to integrate AI into curricula, which could provide a partial solution. Zhou felt that many of the new entrants to the workforce he sees are already pretty tech-savvy. Some are so good at SQL they don’t even have to work in Excel, he said. (Then again, he works at a SaaS company, which likely attracts a techier crowd.)
Instead, the experts said they focus on upskilling existing workforces. Around 50% of finance staff have “basic” technology proficiency, Amanda Joseph-Little, a VP in Gartner’s finance practice, said during a session, meaning they only know the tech needed to do their current jobs. Optimally, she said, all finance staff should be at least “digitally literate” (able to use a wider type of tech than is needed for their day-to-day work), and a majority should be either power users or “citizen digital talent,” able to modify and improve technologies (think Zhou’s SQL whizzes).
Some large organizations said they invest heavily in upskilling programs and giving staff AI practice. At Unilever, everyone from the CFO on down takes part in Digital Foundations, a program which covers topics such as advanced analytics, data visualization and storytelling, and “creating value through digital business models,” Joseph-Little said. EY’s employees practice their AI chops with EYQ, its own private LLM, Ryan said. But Gartner panelists also shared ideas that could work for smaller organizations as well, such as AI demo sessions and Shark Tank-style contests that encourage staff to come up with AI solutions to problems.
Finance’s identity crisis. Ultimately, the finance function of tomorrow is going to look very different. For one thing, the lines between finance and IT are beginning to blur. It’s no longer correct to say that “technology work isn’t finance work” or that tech is “some other work that is temporarily parachuted from IT into finance,” Mallory Barg Bulman, CFO advisory leader at Gartner, said during a keynote. “We need to recognize the view that data integration is finance work, that writing Python code is finance work, that data modeling is finance work.”
But the shift’s a deeper one than simply adopting more tech. We’ve been hearing for a long time that CFOs are becoming less transactional and numbers-oriented, and more strategic. Now that change is happening further down the org chart.
Finance “is transforming from an organization that was primarily focused on processing transactions…to an organization where all that is automated,” Ryan said, and finance professionals “need to understand how to leverage those financial results…to drive insight.” Those, she observes, are “two very different things” with very different expectations.
Finance should no longer think of itself primarily in terms of processes, Gartner Senior Director Clement Christensen said during a keynote. That way, he implied, lies obsolescence. “In our research,” he observed, “we asked controllers, ‘What would you do if the close was completely automated tomorrow?’ And some of them said, ‘Retire.’” (See, controllers do have a sense of humor!)
The shift toward automation and AI also raises serious questions about things like accounting regulations. AI-assisted audits are “uncharted territory,” Omar Choucair, CFO at Trintech, told CFO Brew. The field is going to have to grapple with the “risks and pitfalls of having [AI] agents running through the entire finance and accounting” function, he said. (Pity the PCAOB, which amended auditing standards to account for data analysis just last year—the culmination of a process that started back in 2017.)
It’s a brave new world out there—especially if you’re not a tech whiz or ready to grab a beach chair next to those controllers. But accounting knowledge is still essential, Zhou assured us. “I think it’s pertinent that you have someone with the right kind of experience,” he said. “There’s a big requirement to still review and understand the results” from AI and automation “and translate that for the finance team, and persuade people on the insights you’ve gathered.”
“You want to establish accountability still,” he stated.