We recently ran a workshop with over 100 finance leaders on how to use AI tools in their teams. Before the session, we surveyed them. The number one response, by a distance, was this:
‘We know we need to get our heads around AI. We just don’t know where to begin.’
Its worth paying attention to who was saying it. These aren’t people who are behind because they don’t care. They’re behind because they’re drowning in the day-to-day. The month-end, the board pack, the reconciliations, the forecast that shifted again on Tuesday.
The irony is hard to miss… the thing AI could help with most, the repetitive, time-consuming preparation work, is the very thing that’s stopping them from adopting it.
The shift nobody’s ready for
During the workshop, something landed that I think every CFO needs to hear.
The presenter, Peter Beard (ex-Big Four, on the board of the ICAEW, and someone who helps finance teams implement AI best practice), made a point that stood out. He said the role of finance has always been a function of two things: preparation and review. AI is trying to delete the preparation component, so you’re just left with review.
Think about that for a second.
The reconciliation, the aged debtor analysis, the management accounts build, the variance commentary. AI can do all of it. Not perfectly, not without oversight, but well enough that the preparation side of finance is collapsing from days and hours into minutes.
What’s left is the review. The judgement call. The ‘does this actually make sense?’ The ‘what does this mean for the business, and who needs to hear it?’
It’s like removing the scaffolding from a building. If the structure underneath is solid, you’re fine. If it was the scaffolding holding everything up… you’ve got a problem.
That’s the shift. And most finance teams aren’t structured for it.
When preparation disappears, what’s left?
If you strip away the preparation work, you’re left with the skills that have always mattered but were easy to hide behind ‘busyness’.
Interpretation. Not just producing the numbers, but knowing what they mean and which ones to flag before someone else spots them.
Influence. Getting the commercial team to actually listen when the data says something they don’t want to hear.
Challenge. Pushing back on assumptions in the forecast, not because you enjoy being difficult, but because you know what happens when nobody does.
Translation. Making sure the board pack doesn’t just land accurately, it lands in a way that drives a decision.
These are human skills. They can’t be automated. And they’re the difference between a finance team that’s trusted and one that’s tolerated.
And this is where it gets uncomfortable. If your team has been rewarded for being busy, for producing volume, for grinding through month-end on heroics and late nights, AI doesn’t make that team more efficient. It makes them less relevant. Because the value was never in the preparation. It was in what came after it. And if ‘what came after it’ was just… more preparation, then there’s a problem.
What ‘good’ looks like is changing
This has direct implications for how you hire.
For years, other than the softer skills, finance recruitment has been built around technical competence. Can they close the month? Can they build a model? Have they worked with the right ERP? Those questions still matter. But they’re no longer enough.
If AI can produce the 13-week cash flow forecast in minutes, then ‘can your hire build it?’ stops being the right question. What matters is whether they can interrogate it, spot what’s wrong, and explain to the CEO why the assumptions need to change.
That’s a different skill set. Judgement under pressure. Commercial awareness. Credibility with stakeholders who don’t speak finance. The ability to say ‘the numbers are telling us something’ in a way that actually gets heard.
I’ve written before about credibility risk, the gap between finance being right and finance being listened to. AI makes that gap wider, not narrower. Because if the preparation is automated, the only visible output of your finance team becomes the quality of their thinking. And if that thinking is shallow, or late, or delivered in a way that doesn’t connect… everyone notices.
The shadow AI risk
One thing from the workshop that should make every CFO sit up. Peter flagged that when businesses don’t provide their teams with proper AI tools, people don’t wait. They use personal accounts. They put company data into consumer-grade AI products with no controls, no governance, and no data protection.
Shadow AI is already happening in finance teams. And it’s a bigger risk than most CFOs realise.
Banning AI won’t fix it. Leading the adoption will. Set up the controls. Choose the tools. Define how the team uses them and what the review process looks like. Because if you don’t, your team will figure it out themselves, and you’ll be the last to know.
That, by the way, is a dependency risk and a credibility risk wrapped into one. The very human risks I’ve been writing about for months, showing up in a new costume.
The question worth sitting with
AI will take the parts of the CFO’s job that were never the real job in the first place.
The preparation, the data crunching, the formatting, the manual reconciliation. Those were the things that filled the time. They were never the things that built trust with the board, that protected the business from surprises, or that made the difference between a finance team that’s respected and one that’s just… there.
What AI does is strip all of that away and leave you with a question: when the ‘busyness’ is gone, what’s actually left?
For some teams, the answer is everything that matters. Judgement, influence, and trust.
For others, the answer is less comfortable.
So here’s the question worth asking yourself this week. If AI took over 80% of the preparation work in your finance team tomorrow, would you be confident in what’s left? Would the judgement hold? Would the insight be there? Would the right people be sitting in the right seats?
Because AI won’t tell you if your team is trusted. It’ll just make it much harder to avoid finding out.
Be part of the next conversation…
Click here If you’d like to register for one of our monthly AI and Finance workshops.
