AI is entering finance quickly. CFOs and finance teams are experimenting with ChatGPT, copilots, AI agents and AI-driven analytics. The practical question this session asks is where AI genuinely improves finance work, and where deterministic, reproducible analytics remain the better choice. A mature finance function knows the difference.
Finance has spent decades building controls around data, reporting and decision-making. Teams know where numbers come from, how calculations are performed and how results can be reproduced. Those principles still matter once AI is in the picture. Large language models are probabilistic, so their outputs can vary and it can be harder to trace which fields were used, how data was joined and why a particular conclusion was reached. For many finance decisions, that traceability is exactly what a CFO needs.
One recent study found that 54% of CFOs cannot explain the outcome of their AI-driven analytics.
A five-level maturity model
Level 1 — Experiment: People use AI individually for quick wins like writing, summarising, research and coding. The gains are real, but results vary, confidential data can be exposed, and there is little record of how an output was produced.
Level 2 — Understand: Finance starts to separate different uses of AI and analytics. Drafting an email with an LLM is a different task from asking it to combine datasets and calculate a variance. Teams learn what happens between source data and conclusion.
Level 3 — Control: AI and analytics become part of controlled processes. Data sources are defined, access is managed, and procedures are documented. Outputs can be traced back to source data. The question shifts from "can AI do this?" to "can we show how this result was produced?"
Level 4 — Reproduce & Validate: The same source data and the same logic produce the same result. Data lineage is visible, transformations are documented, and AI-generated scripts and conclusions are validated before finance relies on them. Sometimes the more mature decision is to keep the calculation itself deterministic.
Level 5 — Rely: AI, deterministic analytics and human judgement work together. Automation runs where processes are controlled, AI is used where it adds value, and deterministic analytics handle the work that needs precision and auditability. People stay responsible for challenging conclusions and deciding whether a result is reliable enough to act on. The goal is controlled reliance, not autonomous finance.
Six questions every CFO should be able to answer
Reproducibility: will the same data and logic produce the same outcome tomorrow?
Audit trail: can we reconstruct the steps, filters, calculations and transformations behind the result?
Data lineage: do we know which source systems, files and fields were used?
Explainability: can a finance professional explain why the outcome occurred without saying "the AI found it"?
Data integrity: can we show that joins, mappings, aggregations and exceptions were handled correctly?
Accountability: who owns the outcome when AI produces a plausible but wrong answer?
What you'll take away
A framework for deciding where AI belongs in finance
A clear line between deterministic and AI-driven analytics
Six questions for judging whether an AI-driven analysis can be relied on
A view of maturity measured by how well finance understands, controls, reproduces and explains its numbers, rather than how much work is handed to AI