The research versus development boundary
The boundary is the first date on which an AI build meets all six IAS 38 recognition criteria at once. It sets the earliest date capitalisation can begin. If an entity cannot distinguish the research phase from the development phase, IAS 38 requires the whole expenditure to be treated as research and expensedS1.
Why the boundary matters
The boundary is not a formality. It fixes how much of the programme spend can sit on the balance sheet. Place it too early and the position is not defensible; place it too late and the entity understates the asset. The date must be evidenced, not asserted IAS 38§57.
The method
Work through the four steps in order and record the evidence for each.
- Establish the objective is a specific asset. Confirm the work now targets a defined, identified deliverable, not open exploration. A written specification for the model and its intended internal use marks the shift from research toward development.
- Test technical feasibility. Document evidence that completing the model to its specification is technically feasible, for example a working prototype meeting target metrics on a held-out set.
- Confirm intention, ability and resources. Record management commitment to complete and use the asset, and evidence of the technical, financial and other resources available to finish it.
- Fix the capitalisation start date. The boundary is the first date all six recognition criteria are met simultaneously. Costs from that date forward are capitalisable; earlier costs stay expensed.
The default when you cannot tell
If the phases genuinely cannot be separated, the defensible default is to expense the whole expenditure as research IAS 38§53. This is the conservative position an auditor will expect where contemporaneous evidence of the boundary is thin.
- S1IAS 38 Intangible Assets, IFRS Foundation (IFRS). https://www.ifrs.org/issued-standards/list-of-standards/ias-38-intangible-assets/