CFOs find AI payoff patchy

Bain & Company says chief financial officers are increasing AI spending fast, even though most finance teams still have not turned pilots into scaled results. (bain.com) In Bain’s 2026 survey, 56% of senior finance executives said their companies are increasing enterprise-wide AI investment by more than 15% this year. Over the next two years, 83% plan increases above 15%, and 42% expect increases above 30%. (bain.com) The catch is execution. Bain said only 15% to 25% of CFOs have fully scaled AI in their departments, and just 31% rate AI outcomes in finance as strongly positive. (bain.com) The firms that have scaled it are reporting better results than the firms still testing it. Bain found 41% of companies that scaled AI in finance were satisfied with outcomes, versus 25% of companies still in pilot mode. (bain.com) The early payoff is showing up more in speed than in margin. Bain said 48% of CFOs named speed and cycle-time reduction as their biggest AI win, ahead of headcount or cost savings at 34%. (bain.com) That lines up with a broader complaint now surfacing in finance circles. CFO.com reported April 28 that many companies are investing heavily in AI without measurable gains in EBITDA, cash flow or return on invested capital. (cfo.com) Daniel Schmeltz of Alvarez & Marsal wrote in CFO.com that the problem is less the software than the operating model around it: pilots often lack a financial owner, a baseline, or a defined economic target. (cfo.com) A December 2025 CFO.com report pointed to the same gap from another survey. Only 14% of 200 U.S. finance chiefs surveyed by RGP said AI investments had produced a clear, measurable impact, while 66% said they still expect impact within two years. (cfo.com) Data and systems are a big part of the bottleneck. In that RGP survey, 86% said legacy tools were a significant or moderate barrier to adoption, 35% named data trust as the top barrier to AI ROI, and only 10% said they fully trust their enterprise data. (cfo.com) Gartner is making the case that CFOs should stop treating AI as one project with one return formula. At its March 2026 finance symposium, Gartner said AI should be managed as a portfolio of routine automation, decision-support use cases, and larger transformational bets. (cpapracticeadvisor.com) Gartner pushed that argument further on April 28. CPA Practice Advisor reported that Gartner expects organizations that deploy AI strategically across the technology portfolio to unlock an additional 10 points of margin growth by 2029. (cpapracticeadvisor.com) For now, the numbers suggest finance chiefs are still buying into AI before they can consistently bank the returns. The spending wave is here; the hard part is wiring AI into budgets, controls, data and daily decisions. (bain.com) (cfo.com)