Blog: Finance and AI
Your CFO should help choose the AI problem, not just approve the budget
Published by Forever Foundry
Finance is often invited into an AI initiative when somebody needs a number approved.
By then, the problem may have been defined, the technology selected and the promised return calculated.
The CFO is being asked to validate the arithmetic, not help make the investment decision.
Finance is moving upstream
On 30 September, IBM published research conducted with Oxford Economics among 1,500 CFOs and equivalent finance leaders across 33 geographies and 26 industries.
Of those surveyed, 62% said their role had expanded into technology or AI strategy, while 56% reported greater responsibility for portfolio management and capital reallocation.
Yet only 6% described their finance function as transformation-ready, with AI consistently embedded in workflows and decision-making at scale.
These are self-reported survey findings, not evidence that involving the CFO will automatically produce a return from AI. But they point towards an important change: finance leaders are increasingly expected to shape technology decisions, not simply approve them. Read IBM’s findings and methodology.
The FOFO™ view
The CFO should enter the AI conversation before the business case is written.
Not because finance should own every AI initiative. Because it can help distinguish an operational improvement from a financial claim.
Imagine a manufacturer considering AI to reduce unplanned equipment downtime.
Engineering may understand the faults. Operators may know which warning signs are routinely missed. Production may see the effect on schedules. IT may understand the information and integration required.
Finance can ask a different set of questions.
Does every hour of avoided downtime create additional output? Can that output be sold? Does the improvement reduce overtime, waste or maintenance expenditure? Or does it create capacity that the business is not currently able to use?
Those questions do not weaken the opportunity. They make its value more credible.
Define how the benefit becomes real
“Time saved” is not automatically a financial return.
Neither is “downtime avoided”, “faster processing” or “better productivity”.
Before an AI initiative competes for investment, establish:
- The operational measure expected to change
- Its current baseline and frequency
- How the improvement creates revenue, releases capacity, reduces cost or manages risk
- The implementation and ongoing operating costs
- Who owns the outcome
- When and how the result will be measured
- What other opportunities are competing for the same money and attention
This also helps identify when AI is not the right first investment.
If recurring downtime is caused by a slow approval, missing maintenance information or unclear responsibility between teams, the answer might be process redesign, integration or a low-code workflow. An AI model would not remove the underlying friction by itself.
Cost of change matters as much as cost of build. Read why the fastest AI build can create the longest dependency.
Build the evidence together
Finance should not be handed a finished proposal created from one department’s assumptions.
The FOFO Process™ brings the relevant perspectives into the decision earlier.
AI Compass™ listens across the organisation to identify the business challenges, friction points and opportunities people can see. Slid3rs™ keeps prioritisation human, helping teams compare those opportunities by business value and delivery complexity. FlowRoom™ brings leadership together to agree what deserves investment first.
Finance contributes evidence about how value will be realised. Operations contributes evidence about how the work actually happens. Technology teams help establish what delivery will require.
The result is not simply a better business case. It is a better decision about which opportunity deserves one.
The CFO’s most valuable contribution to AI may not be saying yes or no to a finished proposal.
It may be helping the organisation choose the right problem before anybody starts writing it.
Some of the clearest value sits in work that waits for one expert. Read what stops when your most experienced person is unavailable.
Build the evidence before asking for the budget. Discover which business challenges are worth investing in with AI Compass™.
Related reading
Next