Your Finance Team Is Already at Peak Headcount

Three shifts reshaping finance — whether your organisation is ready or not.

I've spent the better part of the last two decades inside finance functions — as an accountant, a systems architect, and now as someone who helps CFOs use AI to work smarter. I've sat through a lot of keynotes about the future of AI. I've also watched a lot of organisations be caught completely off-guard by shifts they could see coming but weren't prepared for.

So let me be direct with you — peer to peer — about three things I believe every CFO needs to sit with right now.

A9_03169 (1)

Shift 1: Your Finance Team Is the Largest It Will Ever Be

That's not a provocation. It's a reasonable projection based on what AI can already do today, in production, at scale. When I say this to CFO audiences, the room gets quiet. Not because it's alarming — but because most people already sense it's true.

The finance function has expanded over the last decade to absorb complexity: more entities, more reporting requirements, more data sources, more reconciliations. AI doesn't just automate the routine parts of that work — it reasons through it. Agentic AI systems can now reconcile sub-ledgers, propose journal entries, generate variance commentary, and flag anomalies before they become audit findings. These are not prototype capabilities. They are live, in production, today.

The question is not whether AI will reduce headcount in finance. The question is whether you will be the one to shape how that transition happens — or whether it will happen to you.

CFOs leading vs. reacting

Finance teams that actively plan for AI-led efficiency now

"Those who model the transition retain top talent by redeploying them to higher-value work"

Shift 2: The Board Conversation Is Already Happening

Here's something I've been hearing consistently from finance leaders across Australia and New Zealand: the board is already talking about AI productivity. And not in a vague, aspirational way. They're asking specific questions.

If a budget proposal lands on a board table today requesting additional FTE, the first question is almost certainly: how does AI factor into this? The expectation that the finance function will grow to handle more work is being challenged at the governance level. CFOs who can answer that question — with a clear model of what their AI strategy looks like and what it will deliver — are in a very different position to those who can't.

This isn't a future concern. I've had this conversation with CFOs in banking, telecommunications, and manufacturing. The board expectation is live. The CFOs who are prepared for it have already started mapping their AI strategy to their finance operating model. Those who haven't are running behind.

If something goes to the board requesting additional FTE, the first question will be: how is AI playing a part in that proposal? That is now a baseline expectation.

Shift 3: The AI Token Bill Is Coming

This is the one most CFOs are least aware of — and it concerns me the most.

Right now, organisations are deploying AI tools across their teams at speed. Copilots, ChatGPT enterprise, AI assistants embedded in every SaaS platform. The cost of this is largely invisible because it's bundled into software licensing or absorbed into IT budgets. But as AI usage scales — and as agentic AI workflows begin processing thousands of transactions per day — the compute cost becomes a real line item.

Every time your AI assistant answers a question, runs an analysis, or executes a workflow, it consumes tokens. The big players — Microsoft, Google, OpenAI — have spent billions building out that infrastructure. Someone will pay for it. And that someone is us.

The CFOs who are modelling token costs now, building governance frameworks around AI usage, and designing agentic workflows to be compute-efficient — they will avoid a very unpleasant surprise in 12 to 24 months. The others will find a line item on their P&L they didn't budget for.

What This Means for You

These three shifts aren't a reason to slow down your AI adoption. They're a reason to be intentional about it. The finance leaders I work with who are getting the most value from AI right now share a few common traits:

They started with the data. AI on top of a broken or fragmented data foundation doesn't fix the fragmentation — it amplifies it.
They built governance into the design. Not as an afterthought. Every AI-generated proposal in their finance stack requires human approval before it posts.
They're modelling cost alongside value. They understand that AI has a consumption cost, and they're managing it the same way they manage any other input.

The shift is already underway. The question is whether you're shaping it or absorbing it.

In the next blog in this series, I'll get into the specifics of integrated planning — why the absence of it is a liability, and what it looks like when it's working properly with AI infused throughout.

Amendra Pratap is the Founder and Managing Director of Octane Solutions, an IBM Gold Partner specialising in AI-powered finance transformation across Australia, New Zealand, and the Pacific.