Strategy & Value ยท STR-06
Multi-year AI roadmap tied to planning
The organisation has a phased AI roadmap that is linked to its business and financial planning cycles and shows the points where it will decide whether to proceed.
Why it matters
A strategy without a roadmap is hard to fund and hard to hold anyone to. Linking the roadmap to normal planning cycles means AI work competes for resources in the same process as everything else, and explicit decision points stop the organisation drifting into large commitments without choosing to.
Levels
| 0 Absent | Confirmed absent: the evidence shows nothing is in place. Not known is recorded as insufficient evidence, not 0. |
|---|---|
| 1 Initial | No roadmap; AI work is planned project by project, if at all. |
| 2 Developing | A roadmap exists but is a list of projects, is not linked to business or financial planning, or has no decision points. |
| 3 Defined | An approved, owned multi-year roadmap follows the planning cycles, sets decision points with criteria and names the capacity planned uses depend on. |
| 4 Managed | Progress against the roadmap is reported, decisions at each point are recorded with their evidence, and slippage is acted on. |
| 5 Optimising | The roadmap is revised each planning cycle from delivery evidence, value achieved and external change. |
Level 3 criteria
Level 3 is met when every applicable criterion is met. Criteria marked with an overlay apply only to organisations with that characteristic (overlays).
- STR-06.1 An approved, phased AI roadmap covers several years.
- STR-06.2 The roadmap is aligned with the organisation's business and financial planning cycles.
- STR-06.3 The roadmap names its decision points and the criteria for each.
- STR-06.4 The roadmap has a named owner.
- STR-06.5 The roadmap names the capability, data and technology capacity, including compute and storage, that planned uses depend on, and shows which uses depend on which.
Assessment questions
- Is there a roadmap for AI, and how far ahead does it look?
- How does it connect to the organisation's business plan and budget cycle?
- What are the next decision points, and what will each decision be based on?
- The approved roadmap, with phases and decision points
- Business or financial planning documents that reference it
- Records of decisions taken at roadmap decision points
- Roadmap entries for enabling capability, data and compute capacity, with dependencies
How to reach Level 3
- Set out the planned AI work in phases over several years, aligned to the organisation's business and financial planning cycles.
- Name the decision points in the roadmap and the criteria for deciding whether to proceed, change course or stop at each.
- For each phase, list the capability, data and technology capacity (including compute and storage) that the planned uses depend on, and the dependencies between them.
- Name an owner for the roadmap and get it approved through the same route as business plans and budgets.
Common pitfalls
- Producing a list of projects with dates but no decision points, so the roadmap is never questioned.
- Planning uses without the data, skills and compute they depend on.
Starter artefacts
- Phased AI roadmap with decision points and criteria
- Dependency register for capability, data and technology capacity
Cross-references
| ISO/IEC 42001 | 6.2 partial; A.4.2 partial; A.4.5 partial |
|---|---|
| MITRE AI Maturity Model | P2.1 partial |
| UNESCO AI Maturity Framework | 1.3 direct; 3.2 partial |
direct: this control at Level 3 meets the item. partial: it covers part of it.