Dive Brief:
- Finance teams must set realistic time-to-value expectations for their artificial intelligence investments as pressure mounts to deploy the technology more strategically, according to Gartner analysts.
- The research firm’s latest study of AI in finance unveiled Thursday found that 55% of CFOs reported positive overall returns from their 2025 AI initiatives. But when asked about individual AI use cases, 57% said returns were unclear, underscoring the challenge of determining which investments are actually paying off.
- “There’s some projects that are delivering a significant amount of value,” Marco Steecker, a senior director analyst in Gartner’s finance practice, said in an interview. “But a large amount of projects are giving maybe a little bit of productivity” without a significant transformation.
Dive Insight:
Gartner says AI adoption in finance has reached a point where experimentation alone is no longer enough. As finance teams move from pilots toward broader deployment, CFOs need to manage AI initiatives as a portfolio — setting clear expectations for how quickly different projects should deliver value, cutting losses when they fall short and shifting resources toward applications that can have a broader impact on finance and business outcomes.
“The goal is not to stifle experimentation, but to know where to invest, when to cut underperforming initiatives, and which foundational capabilities to accelerate — especially as AI technology becomes more user-friendly and barriers to experimentation diminish,” the report said.
For relatively straightforward finance use cases, including data extraction, accounts payable and receivable automation and report creation, Gartner found that organizations typically achieve expected value within nine to 10 months.
More complex applications, such as data management, insight generation, forecasting and scenario planning, generally require longer development periods.
Those benchmarks can help CFOs distinguish between an initiative that needs more time and one that is consuming resources without a clear path to value, Steecker said.
“If you are investing in an AI tool for data extraction, for accounts payable process automation, or report creation, if you're not seeing value within the first nine months, 10 months of deployment, then you need to go and actually take a hard look at that investment,” he said.
CFOs shouldn't necessarily wait until that deadline to identify problems. Finance leaders should be looking earlier at whether the underlying data is adequate, whether employees are using the technology and whether they understand how to use it effectively, Steecker said.
The challenge is not simply determining when an AI project should produce value. Finance leaders also need to define what value they expect the technology to deliver before an investment begins.
That can be relatively straightforward when AI is used to improve a specific finance process, Steecker said. A dynamic credit-analysis application, for example, could be evaluated based on whether it helps finance teams make faster, higher-quality credit decisions, with an eventual effect on cash flow.
The measurement becomes more difficult when AI is used for broader or more transformative purposes.
Productivity-focused use cases, while seemingly straightforward when it comes to demonstrating AI gains, can present a related challenge, according to Steecker.
“If we're measuring it in terms of X amount of employee time saved, that in and of itself is not an outcome,” he said. “What we need to be measuring is how is that capacity that we've created now being used to deliver on other business outcomes.”
That distinction matters because productivity remains the dominant focus of finance organizations' AI efforts. Improving productivity was the top AI objective reported by finance leaders in Gartner’s survey, with 73% identifying it as a goal in 2025. Cost reduction was second at 59%.
Other objectives, including evolving the enterprise, managing risk, fostering resilience and increasing revenue, were cited by substantially fewer respondents, generally in the 20% to 30% range.
Steecker said finance leaders may be missing some of the broader potential benefits of AI.
“It does feel like finance leaders potentially have some blinders on where AI can really be delivering value for them,” he said.
