76% of finance leaders under high pressure to fully implement AI

Even if there are finance chiefs who don’t see the value of AI or simply don’t have an affinity for AI, most don’t have a choice, as three-quarters of them say they are under high (67%) or very high (9%) pressure to fully implement AI.

And once they implement it, the oversight burden doesn’t disappear: almost all CFOs now spend at least 10% of their workday verifying or correcting AI outputs for finance, and nearly one-third (32%) spend between 26% and 50% of their entire day on the task.

CFOs will likely stay in charge of it, as well: 84% say they are the owner of their organization’s AI budget.

The numbers come from Datarails, an FP&A platform for companies and finance teams that rely heavily on Microsoft Excel. The company hired Global Surveyz Research, an independent firm, which surveyed 270 U.S. CFOs and other finance leaders at companies with at least 1,000 employees and $100 million in annual revenue.

The research found that the surveyed companies spent an average of $361,000 in 2025 on payments or subscriptions for finance and accounting software, and nearly one in 10 spent more than $750,000.

Choosing what to spend that budget on likely won’t be getting any easier. “For CFOs building next year’s budget, the question is whether that spend is building toward AI-ready infrastructure or simply maintaining what’s already there,” Datarails wrote in its survey report.

There is a clear need for disciplined spending, as almost a third (32%) of participants reported that their organizations had exceeded their AI budgets by at least 10% in the 12 months before the survey was taken in July of this year.

Just over half (53%) of finance leaders said they were planning to expand their organization’s AI licenses in the following 12 months, while 40% expected to maintain their current AI spend. Just 7% were looking to decrease such spending.

The top general-purpose AI tool finance teams are currently using is Microsoft Pilot, common to 93% of the survey participants. Datarails called that unsurprising, “given that it ships natively as part of Microsoft 365.” It’s also natural that Perplexity is the least-used such tool, “given that it is seen largely as a research [tool] and has therefore not found significant traction within finance.”

Other survey findings include the following:

• Lack of auditability was found to be the biggest reason for CFOs’ hesitation to fully trust AI with mission-critical tasks.
• A majority of CFOs (60%) said that as AI takes on more finance tasks, they are redeploying staff to higher-value work rather than letting them go.
• Only 23% of CFOs trust AI without human review for revenue forecasting. This drops to 13% for variance analysis, 11% for cash flow management and spend control, 5% for board-ready financial reports and 4% for month-end close tasks.