The CFO operating model — how the finance leader organises people, processes and technology to deliver the function’s remit — is being reshaped by AI faster than most job specifications have caught up with. The change is not that AI does the CFO’s job. It is that AI changes what the finance function spends its time on, and therefore what the CFO needs to lead. As one analysis put it, what AI changes is the scale, speed and complexity at which financial judgment must now operate.

This guide looks at how AI shifts the CFO’s remit, what it does to span of control, how the operating model has to be redesigned so that speed and control move together, and why the finance leader — not the technology function — is the person who owns that redesign.

From processing to judgement

The clearest effect of AI on the finance function is that routine processing compresses. McKinsey found that in functions where AI is robustly adopted, finance professionals spend 20 to 30 percent less time crunching data. Reconciliations, first-draft commentary, data gathering for board packs — these shrink. That does not remove the need for people; it moves the centre of gravity of their work toward judgement, review and interpretation. The CFO’s operating model has to reflect this: fewer hours on production, more on the review and governance that AI-assisted work demands.

This is why the better analyses describe the modern CFO not as someone who adopts technology but as someone who architects decision-making, productivity and governance systems. The task is to redesign how work is done in a hybrid human-and-agent function — deciding where autonomy genuinely reduces friction and where it merely accelerates inefficiency.

The rise of agentic AI changes the model further

The shift underway in 2026 is from AI that drafts to AI that acts. Deloitte’s research shows 54% of CFOs rank integrating AI agents into finance as a top transformation priority, and nearly half of finance leaders with a strategic role have already deployed agents in specific finance activities. Agentic AI allows a function to increase output without a proportional increase in headcount — but McKinsey is clear that this holds only if agents are deployed selectively and governed rigorously. That caveat is the whole operating-model challenge in a sentence.

What happens to span of control

A finance function where AI absorbs routine production can often operate with a flatter structure — but only if the finance leader builds the review discipline to match. The risk is a function that produces more, faster, with less human checking, until an error reaches the board. The CFO’s job is to redesign the operating model so that speed and control move together rather than trading off.

Practical implications for the operating model

Review capacity becomes the constraint, not production capacity. Design roles around who checks and interprets, not only who produces.
Governance moves from an annual policy exercise to a live part of the operating model — who signs off AI-assisted outputs, and on what basis.
Autonomy is granted deliberately. Decide where an agent may act unsupervised, where it may act with review, and where it may only draft.
The finance leader needs a clear line of accountability for AI use, so that a flatter, faster function does not become an ungoverned one.

The CFO as the owner of AI in finance

In most businesses no one has explicitly been made responsible for how AI is used in the finance function. By default that responsibility sits with the CFO, whether or not it has been named. A deliberate operating model makes it explicit — the finance leader owns AI adoption, sets the policy, and is accountable to the board for it. This is a theme we develop in our guides on board-level oversight of AI in finance and building an AI-literate finance function.

Avoiding the common pitfalls

McKinsey’s work on finance functions identifies recurring reasons that AI efforts stall, each of which is really an operating-model failure. Waiting for perfect data before rewiring processes; trying to transform the whole function at once rather than domain by domain; and launching pilots with no roadmap, which rarely scale. A finance leader designing the operating model should build against these directly — deliver value on today’s data, transform in sequence, and anchor every pilot to a defined business need. We cover the sequencing in our guide on building a finance function AI roadmap.

A worked example: what the shift looks like

The operating-model change is easier to see through an example. McKinsey documents a global organisation that used AI-driven contract monitoring to identify value leakage equivalent to roughly 4% of total spend — value that had gone unnoticed despite established controls. The point is not the tool; it is that the gain came from embedding AI directly into an end-to-end financial workflow, where it could act continuously rather than episodically. A traditional operating model checks periodically; an AI-enabled one can monitor constantly. Redesigning the model to take advantage of that is the CFO’s opportunity.

The same logic applies across the function — continuous close, rolling forecasts that update monthly or weekly rather than annually, always-on anomaly detection in payables and receivables. Each is an operating-model change, not merely a tool, and each shifts where the function’s people add value.

The talent dimension

The operating model is also a talent model. Finance functions face ongoing accountant talent shortages, and automating time-consuming manual work is increasingly how leaders both improve productivity and make roles more attractive — the tedious work shrinks and the higher-value work grows. A CFO redesigning the operating model should treat this as a retention and attraction lever, not just an efficiency one: the function that offers analysts genuine business-partnering work, with AI handling the drudgery, is the one that keeps them.

Governing autonomy: the defining question of the new model

As finance functions move from AI that drafts to AI that acts, the operating model’s hardest question becomes where autonomy is permitted. Deloitte’s research shows agentic AI is now a top priority for over half of CFOs, and nearly half of strategically-minded finance leaders have already deployed agents in specific activities. But agents that act unsupervised in a finance function carry risks that a drafting tool does not — a wrong action is executed, not merely proposed.

A deliberate operating model answers this explicitly for each process: may an agent act unsupervised, act with review, or only draft for a human to execute? The answer will differ by materiality — a low-value, high-volume reconciliation may warrant more autonomy than anything touching the statutory accounts or a customer outcome. The CFO who maps this deliberately captures the productivity of autonomy without importing its risk; the one who does not either forgoes the gains or absorbs the risk unknowingly.

The model is never finished

A final point about the operating model in an AI era: it does not settle. The capabilities of the tools change quickly, and a model designed around last year’s tools will be suboptimal against this year’s. The CFO’s task is therefore not to design a fixed structure but to build a function that can keep adapting — one where reviewing and, where sensible, extending AI use is a normal part of how the function runs rather than a periodic upheaval. The operating model, in other words, has to include the capacity to keep changing the operating model.

What this means for finance leadership hiring

A CFO or FD recruited today is increasingly expected to arrive with a view on the operating model AI implies — not to have all the answers, but to understand that the function they lead is changing shape. When you hire at this level, the ability to design and lead a modern, AI-aware finance operating model is a genuine differentiator between candidates, and one FD Capital assesses on senior mandates.

Call 020 3287 9501 or email recruitment@fdcapital.co.uk to discuss a CFO appointment where designing a modern finance operating model is central to the brief.

FD Capital — CFO Recruitment

Fellow of the ICAEW | Placing CFOs who can architect and lead a modern finance function since 2018. We recruit permanent, interim and fractional finance leaders across the UK. 4,600+ network. 160+ placements. Shortlists in 3–7 working days.