IFRS 16 turned lease accounting from a footnote exercise into a full-blown data problem. Every lease — property, vehicles, equipment, even short-term arrangements that cross the threshold — must be captured, classified, and modelled as a right-of-use asset and a lease liability. The math itself is not the hard part. The hard part is the volume: hundreds of contracts in inconsistent formats, discount rates that change, payment schedules with escalations and variable components, and remeasurement triggers that arrive mid-year without warning. Most finance teams still run this in spreadsheets, which means every audit cycle becomes a reconciliation marathon.
The friction compounds in three places. First, data extraction: lease terms are buried in PDFs, scanned amendments, and side letters, and someone has to key them into a model. Second, calculation integrity: a single mis-keyed commencement date or an omitted renewal option throws off the liability, the interest accretion, and the depreciation schedule — and the error is invisible until the auditor finds it. Third, remeasurement: when a lease is modified, an index changes, or a renewal option is exercised, the schedule must be rebuilt and the journal entries re-derived. Under time pressure, these steps get rushed, and rushed lease accounting is how restatements happen.
An AI assistant does not replace the accounting judgement — it removes the mechanical load around it. Used properly, it reads lease documents, extracts the terms into a structured register, applies the IFRS 16 calculation logic consistently, produces the amortisation schedules and journal entries, and flags the judgement calls for a human to decide. The controller stays in control; the spreadsheet drudgery and the keying errors go away.
Where AI fits in the lease accounting workflow
The highest-value use is not “calculate my leases” in the abstract — it is a repeatable pipeline: extract, structure, calculate, reconcile, document. The prompts below are built for that pipeline. The first turns raw lease documents into a clean, auditable lease register. The second turns that register into the calculation schedules, journal entries, and disclosure inputs, with every assumption surfaced rather than buried.
One caution before you start: give the model your accounting policy first. Incremental borrowing rate methodology, the short-term and low-value exemption thresholds, how you treat variable payments and non-lease components — these are your standards, and the model should apply them, not invent its own.
First, read these files completely before responding:
[lease_policy.md] — our IFRS 16 accounting policy, including IBR methodology, exemption thresholds, and treatment of variable and non-lease components
[lease_documents.md] — extracted text of the lease contracts and amendments, one section per lease
[register_template.md] — the column structure and naming conventions for our lease register
Here is a reference for what I want to achieve:
[Upload a completed register row for one lease as markdown, or describe the level of detail required]
Here’s what makes this reference work:
Each field is populated from a specific clause in the contract, with the clause reference noted. Judgement items (renewal options, purchase options, variable payments) are flagged rather than assumed. Dates are in ISO format, amounts in [currency], and no field is left blank without a stated reason.
Here’s what I need for my version / SUCCESS BRIEF:
Type of output + length: A markdown table of lease register rows, one row per lease, plus a short exceptions list.
Recipient’s reaction: Our auditors should be able to trace every figure back to a clause without asking us a single question.
Does NOT sound like: A summary of the leases, or a narrative description. No invented terms, no silent assumptions.
Success means: Every lease in the batch has a complete register row, and all judgement items are listed for controller review.
My context file contains my standards, constraints, audience. Read it fully before starting.
DO NOT start executing yet. Ask clarifying questions first.
Give me your execution plan (5 steps max) before you begin.
Notice what the first prompt does not ask for: calculations. Extraction and calculation are separated deliberately, because errors in one should not contaminate the other. Once the register is reviewed and signed off, the second prompt takes it the rest of the way — and because the register is structured, the calculation step becomes deterministic rather than interpretive.
First, read these files completely before responding:
[lease_register.md] — the approved lease register, one row per lease, with all terms and clause references
[lease_policy.md] — our IFRS 16 policy, including IBR methodology, depreciation conventions, and remeasurement triggers
[chart_of_accounts.md] — account codes and naming for ROU assets, lease liabilities, interest, and depreciation
Here is a reference for what I want to achieve:
[Upload a prior-period amortisation schedule and journal entry extract as markdown, or describe the exact output format]
Here’s what makes this reference work:
Liability is measured as the present value of remaining payments at the IBR, with interest accreted using the effective rate and payments reducing the liability. ROU asset depreciates on a straight-line basis over the lease term. Each period shows opening balance, interest, payment, closing balance, and the paired journal entry with account codes. Remeasurement events are shown as separate, clearly dated entries.
Here’s what I need for my version / SUCCESS BRIEF:
Type of output + length: One amortisation schedule per lease for the full lease term, a journal entry summary by period, and a disclosure input table for the year-end note.
Recipient’s reaction: The schedules should be audit-ready on first review, with every remeasurement traceable to a register entry.
Does NOT sound like: A high-level estimate or a single blended schedule. No rounding drift, no unexplained adjustments.
Success means: Schedules tie to the register, journal entries balance, and the disclosure table reconciles to the general ledger balances.
My context file contains my standards, constraints, audience. Read it fully before starting.
DO NOT start executing yet. Ask clarifying questions first.
Give me your execution plan (5 steps max) before you begin.
The practical tip that matters most: run the second prompt on a small batch first — five leases with at least one modification and one variable-payment component — and reconcile the output against your existing spreadsheet model. If the numbers tie, you have validated both the prompt and your policy file, and you can scale to the full portfolio with confidence. If they do not tie, the difference will almost always point to an ambiguity in your policy, which is worth finding before the auditors do.
What to try next: build a remeasurement-only prompt that takes a single modification event and produces the revised schedule and catch-up journal entry, then add an annual disclosure pack prompt that pulls the note inputs directly from the schedules. Once extraction, calculation, and disclosure all run from the same register, IFRS 16 stops being a year-end scramble and becomes a controlled monthly process.
Published on 26 October 2026 on growwithgpt.com