The most important question in a board meeting is often the simplest. Where did this number come from?
For years, finance teams could answer it confidently. A report was built by analysts, reviewed by managers and signed off through a clear process. The work could be traced back to the people who produced it.
Today, that is changing. Board packs increasingly include figures generated by automated workflows, forecasts produced by models and commentary drafted by AI tools. Reporting is faster than ever, but many organisations have lost something along the way. The ability to explain exactly how an answer was produced.
And the errors have started arriving. One in four executives now report audit-detected AI errors reaching boards or external audiences, built on data most of them admit they do not trust (the verification gap). Audit committees have noticed. They are no longer interested only in whether a number looks reasonable. They want to know where it came from, what data it relied on and whether the result can be reproduced.
Those are not technical questions. They are governance questions.
The challenge is not AI. It is traceability.
Most finance leaders are not worried about AI generating more information. They are worried about defending it.
Data flows through multiple systems. Models transform it. AI tools summarise it. Reports are assembled automatically. Somewhere along the way, the connection between a final figure and the steps that created it becomes difficult to see.
When that happens, confidence starts to replace evidence. That is a risk. A number should not be trusted because people believe it is correct. It should be trusted because the organisation can demonstrate why it is correct.
More reviews will not solve it
The usual response is to add controls. Another review meeting, another sign-off, another layer of approval.
But reviews are only as effective as the information available to the reviewer. If nobody can easily show the source data, the transformation steps and the models involved, additional reviews do little more than increase cost and effort.
The answer is not more process. Every number should arrive at the board with its own evidence attached.
What a defensible number looks like
A finance team should be able to answer four questions about any figure in a board pack, immediately.
- What data produced it? The source systems should be recorded automatically as the report is created, not reconstructed later when questions are asked.
- Can it be reproduced? The same inputs through the same process should give the same result, every time. A number that cannot be reproduced cannot be properly investigated.
- What models contributed to it? Where AI or analytical models play a part, the organisation should know which model, which version, who approved it and how it has been performing, in a standing record per model (the Model Passport).
- Did it follow the approved route? Every figure should travel the same controlled path to the board. The greatest risks rarely come from the official process. They come from the spreadsheet emailed the night before the meeting.
Trust comes from evidence
The benefits go beyond compliance. Finance teams spend a remarkable amount of time reconciling reports, checking versions and explaining differences between numbers. Much of that effort exists because the reporting process has no built-in traceability. When every figure carries its own audit trail, scrutiny becomes cheap to satisfy. That is what trust looks like in practice.
When a director asks where a number came from, the answer should take seconds. Not days.
A simple test
Take your most recent board pack. Pick a number at random. Ask how long it would take to show where it came from, what data produced it, how it was calculated, and whether you could recreate it exactly.
If the answer embarrasses you, that is the gap the Prophesee Finance Suite closes. Every figure carries its evidence, regenerates identically on demand, and names the models behind it, because the record is produced by the system as it works, not assembled for the meeting. The ability to generate information is no longer the advantage. The ability to defend it is. Make the pack defensible.