World Model Readiness
Engraved team-functions instrument

For your team

Teams by Function

Module · When AI touches the numbers

The Finance Team AI Check

Finance runs on numbers that have to be right, and AI produces numbers that are only usually right. A model that fabricates a plausible figure with total confidence is a specific hazard when the output becomes a forecast, a reconciliation, or a line in the close. This module checks the five habits that keep AI a help rather than a hidden error: whether your team verifies before trusting, whether AI outputs get reconciled to a source, whether you can audit how a number was produced, whether the same person builds and approves, and how far AI has crept into the close itself.

Question 1 of 5 · Verify before trusting

Does your team verify AI-produced numbers before using them, or trust them by default?

A language model will state a wrong figure as confidently as a right one, with no tell in the output. In finance that confidence is the danger: a fabricated number reads exactly like a correct one until someone traces it. The habit of checking before trusting is the whole defence, and it erodes the moment the tool seems reliable.

Question 2 of 5 · Reconciled to a source

Are AI outputs reconciled back to a system of record, not just accepted as given?

An AI can summarise a ledger, categorise transactions, or draft a reconciliation, and be subtly off in ways that only surface against the source. Reconciliation is the control that catches this: the AI proposes, the system of record disposes. Skip it and the AI's version quietly becomes the truth nobody checked.

Question 3 of 5 · The number is auditable

For an AI-assisted figure, can you show how it was produced, from prompt to formula?

When an auditor or your CFO asks how a number was arrived at, "the AI worked it out" is not an answer. You need the trail: the inputs, the prompt or formula, the model, the version. A figure you cannot reconstruct is a figure you cannot defend, and finance lives on defensible figures.

Question 4 of 5 · Duties stay separated

Does AI let one person now do what used to need two, collapsing your segregation of duties?

Segregation of duties exists so the person who creates a transaction is not the person who approves it. AI quietly erodes this: a single analyst with a capable assistant can now build, populate and approve work that once required a second pair of eyes. The control did not fail loudly, it just stopped applying.

Question 5 of 5 · Close-cycle AI is governed

How far has AI moved into your period close, and is that use controlled?

The close is the highest-stakes, most time-pressured work finance does, which makes it exactly where an unverified AI shortcut is most tempting and most dangerous. Using AI to draft journal narratives is one thing; letting it propose entries or accruals under deadline pressure is another. The question is whether that line is drawn deliberately.

For the statistics · one click each

Three questions for the public picture

These do not affect your score. They feed the anonymised, aggregated statistics; groups under 8 respondents are never shown.

Where does your finance team use AI on the numbers today?

Not on the numbers
Analysis and drafting
Reconciliation and categorisation
Forecasting and reporting
Inside the period close

What does your team do before trusting an AI-produced number?

Trust it as given
A quick glance
Verify when it matters
Verify against a source
No AI on the numbers

Can you reconstruct how an AI-assisted figure was produced?

No trail
Only from memory
Partly documented
Fully reproducible
No AI-assisted figures

Your context

Used to calibrate the report. Company size and sector remain in the anonymized dataset; your email does not.

What the five levels look like

Every dimension in this assessment is scored 1 to 5. This is what the levels mean, dimension by dimension. The graded report diagnoses where your own answers land and what to do about it.

Verify before trusting

  1. 1Trusted blindly
  2. 2Spot-checked rarely
  3. 3Checked when it matters
  4. 4Verified before use
  5. 5Verified and documented

At the low end: Numbers taken from an AI without a check will eventually put a fabricated figure into a decision. Make verification against a source the default before any AI output is used, not an afterthought when something looks wrong. What good looks like: Verifying before use and recording that you did is what lets finance use AI without inheriting its errors. Keep the discipline as the tools improve; trust built on a good month is how the unchecked error gets in.

Reconciled to a source

  1. 1Never reconciled
  2. 2Taken at face value
  3. 3Reconciled sometimes
  4. 4Reconciled to source
  5. 5Reconciled and signed off

At the low end: AI output that never meets the system of record can drift from reality without anyone noticing. Reconcile every AI-touched figure back to the ledger before it feeds a report. What good looks like: Reconciliation to a system of record with a sign-off is the control that lets AI accelerate the work without owning the truth. Keep the source authoritative; if the AI output starts feeding the record it was meant to check, the control is gone.

The number is auditable

  1. 1No trail
  2. 2In someone's head
  3. 3Partial notes
  4. 4Inputs and method saved
  5. 5Full reproducible trail

At the low end: A number with no record of how the AI produced it cannot be defended when someone asks. Start capturing the prompt, inputs and model version alongside any AI-assisted figure. What good looks like: A full, reproducible trail from prompt to figure is what makes AI-assisted finance auditable. Keep it version-stamped; the same prompt on a new model can return a different number, and you will need to show which one you used.

Duties stay separated

  1. 1One person throughout
  2. 2Separation on paper
  3. 3Separated for big items
  4. 4Enforced separation
  5. 5Enforced, AI-aware

At the low end: If one person plus an AI can now build and approve the same work, the check that catches errors and fraud is gone. Restore a second human approver on anything AI helped produce. What good looks like: Enforced separation that accounts for what AI now enables keeps the create-and-approve boundary real. Revisit it as the assistants get more capable; each capability jump is a chance for one person to quietly reclaim both roles.

Close-cycle AI is governed

  1. 1Ungoverned in the close
  2. 2Creeping in unmanaged
  3. 3Used, informally
  4. 4Used with controls
  5. 5Controlled and reviewed

At the low end: AI making its way into the close with no controls puts unverified figures into your most consequential process under the most pressure. Draw an explicit line now for what AI may and may not do in the close. What good looks like: Governed AI use in the close, with the high-risk steps kept under human control, lets you gain the speed without betting the numbers. Review the boundary each cycle; deadline pressure is a constant push to let the AI do just a little more.