AI was everywhere at the 2025 Gartner CFO & Finance Executive Conference. No fewer than 68 sessions were specifically devoted to the topic. Practically every SaaS vendor touted their products’ AI capabilities. The keynote speakers asserted that 60% of finance organizations will use “AI-enabled scenario planning” by 2028, and made ominous pronouncements, such as “There is no future for good process executors. There is still a future for humans, but only humans who want to work on complex, novel problems.” (Outside of Star Trek and role-playing games, you will never be referred to as a “human” as often as you will at a finance conference.)
But Gartner’s also known for its Hype Cycle, fittingly enough, and it raises the question: How much of the AI buzz was just that: hype? Is it really going to transform the finance function for good, or is it merely an extension of technologies that came before it?
As it turns out, the answer’s a little of both.
Like RPA, but more accurate: AI can be divided into “two major buckets,” Deirdre Ryan, global finance transformation leader at EY, told CFO Brew. “One is productivity.” Used in that sense, AI’s like automation on steroids. It’s more accurate than what’s often referred to as robotic process automation (RPA), and it can save finance teams tons of time.
For instance, RPA was able to “read” PDF invoices, extract the relevant data, and match it to a purchase order, Erik Zhou, chief accounting officer at payments platform Brex, told CFO Brew. But it only had about a 70% “hit rate,” meaning around a third of invoices would still have to be read manually. AI, he said, increases that hit rate to around 95%. “You have every single invoice being reviewed automatically by a human-like tool,” Zhou said.
Data mining and text generation take AI beyond automation: But AI’s able to do more than just automate with greater accuracy. The “game changer,” Ryan said, is its “ability to mine” far vaster amounts of data than humans (see, they’ve got me doing it now) can. That allows it to make forecasts, she said.
News built for finance pros
CFO Brew helps finance pros navigate their roles with insights into risk management, compliance, and strategy through our newsletter, virtual events, and digital guides.
Brex’s finance team, for instance, uses AI for “flux analysis,” Zhou said. Finance staff can ask the AI a question or click on a button and see trend lines for different vendors over the next 12 months. “It’s pretty magical,” he noted.
Organizations are also learning to layer “traditional AI,” which is math-based, with generative AI, which is text-based, Gartner director analyst Alex Levine said during a session. Users work with both types in concert with RPA to use each to its best advantage. For instance, during the AP process RPA can, among other things, “ingest and route invoices to a processing system,” generative AI can “extract key terms” from invoices and draft emails to vendors, and traditional AI can identify duplicates or “unusual patterns,” he said.
A newer term, agentic AI, refers to systems that “can also plan and execute autonomous actions, dynamically adapting their approach based on context and ongoing analysis to achieve a specific goal,” Levine said. Several examples of agentic AI were on display at Gartner. Business planning software company Pigment now incorporates AI agents, or tools, that help users unearth insights from their data, Jay Peir, interim CFO and head of strategy, told CFO Brew. For instance, an agent could scroll through a travel report, find that one team is spending more on hotels than others, and suggest reasons why. Customers can then use a ChatGPT-like natural language feature to ask questions about the ideas the agent came up with.
But what are finance teams doing with all the time they save? And, with so much being automated, what’s going to happen to finance jobs? We will explore those questions in part two of this article series.