If you manage leases under IFRS 16, you know how time-consuming it can be to ensure all inputs for a lease liability calculation are accurately identified. That’s where this Python-based “IFRS 16 Lease Contract Reader” comes in. It uses OpenAI’s ChatGPT model to read your contract in Word format and automatically pull out essential data for the lease liability and right-of-use (RoU) asset schedules.
First, you place need to create on your desktop the following folders, and place your lease contract in the Input folder. Please put there only one lease contract that you want ChatGPT to analyse:
Second, you need to update the code to use your OpenAI API key:
Next, the program calls ChatGPT to analyze the lease text. It extracts key variables such as lease_payment, payment frequency, whether payments are upfront or in arrears, total number of payments, and more. It even checks details about direct costs, incentives, and indexation—an often overlooked complication.
A second “review” pass asks ChatGPT to confirm or correct any values. This helps identify uncertainties and offers best guesses if the contract language is ambiguous. Finally, the script writes two things to Excel: a short “Lease Summary” in natural language and a “Reviewed Values” sheet containing the finalized numeric fields. These are precisely the inputs you need to feed into your IFRS 16 calculations, or into another Python tool that calculates the liability and RoU asset schedules, such as the one on our website.
The output looks as follows:
The result is a faster: you go from raw contract text to a structured dataset, complete with a short summary the accountant can read at a glance. However, properly review the final output, provide an annual discount rate, and use the values to build your lease amortization schedules. It’s a straightforward way to incorporate AI into your IFRS 16 workflow.
Code Download: