Most of what gets said about AI in finance is about what it can take over: reconciliations, first drafts of commentary, variance explanations, the monthly assembly of numbers from too many files. Much of that is real. I build parts of my own tools this way now, and the routine shrinks faster than I expected.
The more useful question comes after that. When routine work shrinks, the hours it used to take do not stay free. Someone decides where they go. In most finance teams that decision is never made on purpose. Whatever arrives first takes them.
Last week's issue described a person planned three times: a day job and two projects, well over 100 percent combined. Free up a fifth of her day job and watch what happens. The day job will not shrink to match. It absorbs the hours, because the day job is the only claim that never has to argue for them.
That is why time saved in finance so rarely shows up anywhere. The savings are real. Their destination has no owner.
So before the tool choice, settle two things: who decides where the freed capacity goes, and against which priority. That is a steering decision. It belongs on the same agenda as the software, and ideally earlier.
Over the coming weeks, The CFO Lens looks at AI from this angle. Less about what it can do, more about what it changes in who decides.
If the routine work in your team halved tomorrow, who would decide where the hours go?
Cordula Buss · Plan A2C / Helping finance and programme leaders build steering logic that works.