By now, every strong finance leader knows AI matters.
Many are already testing tools and seeing where AI could change how finance work gets done. Turning that early progress into a board-level conversation is the harder part.
The challenge is that AI can be hard to pitch at the right level. Go too deep into tools, prompts and models, and the conversation becomes too technical. Stay too high-level, and it can sound like another set of experiments with no clear return.
The job for finance leaders is to translate AI into something the board can actually work with: where the value is, how risk is being controlled, what capability needs building and what decision is needed next.
Don’t confuse AI activity with AI progress
A mistake I often see is finance leaders reporting on AI activity when the board is really looking for progress. Early experiments create momentum, build confidence and help teams see what’s possible. But a list of pilots isn’t the same as a business case.
What the board is looking for is clarity on which finance processes have improved, what value has been measured and what they need to do about it.
So the work needs to start with the process, not the tool.
Before launching pilots, finance leaders should be clear on the specific problem they’re trying to solve. The use case should be narrow enough to test properly, with human review in place and a simple framework for sign-off and risk.
That gives you a stronger board message: “We’ve used AI to improve this part of the forecast process. Here’s what changed, here’s how we controlled the risk, and here’s the decision we need from the board.”
Speak in the board’s language
The board doesn’t need every detail of the use case. Especially when talking to non-finance stakeholders, it’s important to use language that actually resonates.
AI creates decisions around investment, risk appetite, governance and capability. Those are board-level topics. If finance leaders want support, they need to frame the conversation around the outcomes the board is responsible for.
That requires moving beyond what AI does inside the finance team and explaining what it changes for the wider business.
A simple way to land this is to use a one-sentence headline, then explain the “so what”. For example:
- “We’re seeing the clearest AI return in speed, not yet in cost reduction. That means we should measure this phase through cycle time, review quality and adoption, rather than forcing a premature headcount saving.”
- “We’re seeing informal AI use in the team. That means we need clearer approved routes before usage moves further outside our control.”
The idea is simple: make the point, explain why it matters, then pause. At board level, clarity often beats completeness.
Be careful how you talk about ROI
Boards will undoubtedly talk about ROI, and this is another place finance leaders trip up.
As I’ve written about before, most finance AI projects struggle to show ROI because they’re measuring the wrong thing. They look too quickly for headcount reduction or full automation, when the technology just isn’t there yet.
Capability is often the strongest early use case. AI can help teams work faster, but the bigger opportunity is improving how finance work gets done.
I’d frame the value across four areas:
- Speed: work gets done faster with more time for higher-value tasks.
- Accuracy: outputs become more consistent and easier to review.
- Risk reduction: more work moves into approved, controlled workflows.
- Economics: finance creates more capacity for the same or lower cost.
At this stage, the board shouldn’t only be looking for direct cost reduction. It should also be looking for better processes, stronger habits and evidence that the team is building the skills it will need next.
There’s still a cost argument. If 35 people have AI licences at £600 each, that’s £21,000 a year. In many finance functions, that’s a relatively small investment if it helps people improve processes and build capability.
Talk about risk before it becomes an objection
Risk shouldn’t be hidden at the end of the conversation. Raising it early builds trust and shows the opportunity is being taken seriously, but not casually.
The main concerns are familiar: sensitive data, inaccurate outputs, unclear ownership, weak review and people using tools outside approved routes.
Importantly, the risk conversation needs to be more practical than “stop people using AI”.
A stronger board message would be: “We know AI is entering the way work gets done. Our job is to make safe use easier than unsafe use.”
For finance, the next step is to put simple guardrails around real workflows: approved tools, clear data rules, human review where judgement is involved, and sign-off before a use case is scaled.
The board needs to hear that finance is moving forward carefully, with enough control to build confidence.
Keep human judgement in the story
One concern boards have is whether AI weakens judgement or creates over-reliance.
Finance leaders should be clear that, for now, the best use cases are usually about improving how people work rather than removing people from the process.
If AI is presented as a replacement for judgement, the board will naturally worry about risk and accountability. If it’s presented as a way to improve preparation, consistency and review, the conversation becomes much more credible.
AI can help draft commentary, suggest forecast questions or summarise information. Finance still owns the judgement. That is the line board members need to hear clearly.
Create confidence by ending with a clear decision
Finance leaders need to be clear on what they want from the board. That might be approval to scale a use case, investment in licences or training, support for a clearer AI policy, or agreement on where human review must remain in place.
A useful structure is:
- What have we learned?
- Why does it matter?
- What needs managing?
- What decision do we need?
The aim is to create confidence that AI is being used against real business problems, that the risks are understood and that finance knows what needs to happen next.
If the board leaves interested but unclear on the next step, the conversation hasn’t done its job.
Ready to talk about AI with more confidence at board level?
If you’re a CFO, FD or finance leader trying to turn AI activity into a credible board-level story, this is exactly the work I support clients with. Get in touch for a chat.
