CFOs are funding AI from core budgets, not experiments. Enterprise generative AI spending reached 37 billion dollars in 2025, roughly tripling year over year, while the share funded from experimental budgets fell from 25 percent to 7 percent. The buying pattern behind those numbers comes through clearly in Open Future Forum's first-party research with finance leaders.
Where CFOs are increasing AI spending
The spending shift is structural. When AI moves from experimental to core budgets, it starts competing with everything else the CFO funds, and it is winning those comparisons. Within enterprise generative AI application spending, software development leads at roughly 4 billion dollars for coding tools, followed by the functions where output is measurable: customer support, finance operations and sales. The CFO AI Leverage Report tracks this movement edition over edition.
Which finance workflows automate first
Inside the finance function itself, automation follows the paper trail. The workflows moving first are the ones with high document volume and clear rules: accounts payable and receivable, expense processing, reconciliation and close, and first-draft reporting. Planning and analysis follows, with AI compressing the time between question and answer rather than replacing judgment. The CFO AI Market Map organizes this into eight CFO workflow categories, from record-to-report through FP&A to the emerging agent layer.
Who controls AI budgets
This is the finding with the biggest strategic consequence. In Open Future Forum's early data, three in five finance leaders pointed to the CFO or finance as the AI sign-off seat. Yet when sixteen AI founders were asked who their buyer was, not one named finance. Founders sell to CIOs, CTOs and business units; the person who actually approves the spend is increasingly the CFO. That gap means vendors are pitching the wrong room, and it means CFOs are evaluating AI purchases the sellers never designed for a finance audience.
What CFOs are cutting
The clearest cut shows up in plans rather than line items. Finance leaders' headcount growth expectations fell from 6 percent for 2025 to 2 percent for 2026. The operating model behind AI budgets is growth without proportional hiring: the roles that would have been added to absorb volume are the first thing AI spending replaces. Alongside that, experimental AI pilots without owners are being consolidated into fewer, funded deployments, which is exactly what you would expect once the money moves into core budgets.
How CFOs measure AI ROI
The measurement question has a sharper answer than a year ago. The metric that recurs in Open Future Forum's finance rooms is leverage: can the team produce more output, close faster, cover more entities, without adding headcount or vendor spend in proportion? That is the logic behind the CFO AI Leverage Index tracked in the report series. Cost-per-task comparisons and cycle-time deltas support it, but the headline test is organizational: did the function grow its output while its cost line stayed flat?
The market map behind the answers
These findings come from first-party research: 421 survey responses collected across seven Open Future Forum events, feeding the CFO AI Leverage Report and the CFO AI Market Map with its 64 vendors across eight workflow categories. Early findings with small response bases are labeled directional in the reports themselves.
These same questions are discussed live, off the record, inside the Open Future Forum CFO Executive Forum, where much of this data originates. If you hold the finance seat, that room is where the next edition's answers are forming.
Last updated: August 13, 2026
Frequently Asked Questions
Read the Research, Then Join the Room
The CFO AI Leverage Report and CFO AI Market Map are built on first-party survey data from finance leaders at Open Future Forum events.