Boards are asking CFOs what AI will deliver. The close takes longer than it should. Forecast refreshes lag the business. Finance still runs on spreadsheets. Although investment in AI is rising quickly, execution is slower.

But something is shifting in the CFO suite. The function that has been approving AI budgets for everyone else while remaining skeptical of the technology’s value closer to home is beginning to move.

The capital commitment to AI is real and growing, and finance is catching up. A recent Bain & Company survey of senior finance executives shows 56% are increasing enterprise-wide AI investment by more than 15% this year. Over the next two years, 83% of CFOs plan AI budget increases above 15%, with 42% expecting increases above 30% (see Figure 1).

Figure 1

A significant share of that spending will be allocated to finance, and CFOs see the pull-through: In a concurrent survey of 264 finance department heads, about 75% expect AI budgets in their area to rise, while 22% expect a substantial jump. This is not incremental experimentation; it’s a serious commitment to AI in the operations of a function known for fiscal discipline.

The shift matters because it has not been easy to make. Results from early AI investments have been mixed, with only 31% of CFOs rating AI outcomes in finance as strongly positive. CFOs are doubling down not because early returns have been spectacular, but because the gap between those who have scaled AI and those who haven’t is becoming too large to ignore.

Speed is the first dividend

When CFOs describe their biggest AI win, speed and cycle-time reduction leads at 48%, ahead of headcount or cost savings at 34% (see Figure 2). That ordering matters. Tighter close cycles, streamlined reconciliations, and early variance insight improve a company’s ability to detect exceptions, correct course, and redeploy capital faster.

Figure 2

The strategic stakes are higher than the metrics suggest. In an environment defined by shifting trade policy, volatile rates, and supply chain disruption, the finance function’s ability to reforecast quickly, reallocate capital on short notice, and surface risk in real time is a competitive advantage, not just an operational one. A finance function that compresses the cycle from market signal to management decision from weeks to days is better positioned to help the business move faster than its competitors.

The implication: Redesign the business case and the scorecard. Time-to-insight and time-to-action should sit alongside cost as primary outcomes—measuring days-to-close, forecast refresh cadence, and time-to-variance resolution with the same rigor as headcount and expense.

The scale imperative

The winners won’t be the companies that use AI; they’ll be the ones that scale it. Yet despite rising investment, roughly 60% of finance organizations report AI initiatives still in pilot or limited production. Only 15%–25% have scaled machine learning or generative AI into full production across finance.

The satisfaction data we collected makes the case for scaling more powerfully than any theoretical argument. Among all CFOs, 31% rate AI outcomes as strongly positive. Among those who have scaled any type of AI (machine learning, GenAI, or agentic) into full production, that figure rises to 41%, compared with 25% among those still in pilot mode. Among top-quartile organizations by AI maturity, it exceeds 60% (see Figure 3). The return on AI investment is not primarily a function of how much you spend but of how far you scale. For CFOs seeking broader organizational support, this is the data worth sharing.

Figure 3

Sequencing matters too. Results to date are strongest in transactional finance, especially invoice to cash and procure to pay, even as near-term investment attention shifts toward financial planning and analysis (FP&A) and financial reporting (see Figures 4 and 5). The pragmatic path is to first industrialize value streams where the economics are proven, then expand with a mature scaling engine in place.

Figure 4

Figure 5

Overcoming “workflow debt”

The primary constraint on AI’s next wave in finance is no longer technology. It’s work design, controls, and adoption. As AI maturity rises, organizations increasingly cite change management and talent—not technology—as the principal barriers to value.

The reasons are visible in how finance organizations are actually using AI. Bain’s research finds that roughly 12% of finance organizations have deployed machine learning in FP&A forecasting at full scale—in active production, not piloting or testing. Yet in many cases, the underlying process hasn’t changed. Finance teams run AI-generated forecasts alongside existing bottom-up planning cycles: two processes running in parallel, neither fully trusted, with the expected benefits (faster cycle times, fewer people hours, sharper accuracy) largely unrealized. The AI was deployed. The work wasn’t redesigned.

This is the essence of “workflow debt.” It isn’t a technology problem. It’s what happens when AI gets layered on top of existing ways of working instead of providing the impetus to change them. If workflow debt isn’t addressed, AI and automation can multiply complexity instead of productivity.

The organizations getting the most from AI use it as a forcing function to ask which steps, handoffs, and approvals are necessary. For CFOs with responsibility for ERP or enterprise technology, this is especially relevant: Accumulated process complexity compounds the technical debt already embedded in legacy systems. Workflow debt and technical debt are not separate problems; they intensify each other.

Implications for CFOs

CFOs aiming to convert AI investment into a structural performance advantage need to act on several fronts simultaneously.

Treat speed as a strategic outcome. Redefine the AI business case to center time-to-insight and time-to-action alongside cost. Build a measurement scorecard that tracks days-to-close, forecast cadence, and time-to-variance resolution. When speed is the headline metric, the true value of AI—better decisions, faster course correction, stronger financial control—becomes visible and defensible to the board and business leaders.

Build a scaling engine, not a pilot portfolio. Consolidate experiments into production-ready patterns across data readiness, integration, controls, and adoption. Start where the economics are most proven—invoice to cash, procure to pay, and the accounting close—and expand into FP&A and reporting, which many CFOs are already targeting for investment.

Pay down workflow debt before deploying agents. Map the handoffs, approvals, and exception paths in high-priority processes. Simplify decision rights and stabilize rules before introducing autonomous execution. Finance leaders who do this work first will find that agentic AI deploys faster, performs better, and sustains trust.

Don’t let yesterday’s pilots define tomorrow’s ambitions. AI capabilities have advanced rapidly in the past 12 to 18 months. Capabilities that were unreliable or simply unavailable a year ago are now in production at leading organizations. Companies that benchmark expectations against first-generation tools risk misreading the current opportunity.

CFOs who act on these priorities will move finance from incremental automation to structural performance advantage with faster cycles, stronger governance, and better business decisions. The competitive divide is forming now. The question is which side of it each organization chooses to be on.