See your AI maturity.
Build the roadmap.
AIMM is a structured framework for assessing how mature your organisation's AI capability is — across five dimensions, against five progressive levels — and for designing a sequenced roadmap from any current position to a higher one.
Why maturity models work
Forty years of enterprise precedent.
CMMI for software engineering. ITIL for service management. COBIT for governance. AIMM follows the same proven pattern — applied to the discipline of artificial intelligence.
Common Language
A shared vocabulary for leaders, teams, and partners — without vendor jargon.
Diagnostic, Not Judgemental
Level 2 is data. ‘Immature’ is judgement. One opens conversation, the other closes it.
Roadmap Discipline
Investment sequenced against measurable criteria — not opinion.
Multi-Dimensional Honesty
Be Defined on Data, Initial on Governance, Developing on People — and design accordingly.
What AIMM is
Built for the boardroom.
Evidence-Based
Every level on every dimension is anchored to observable evidence — documents, systems, behaviours, decisions. Not self-rating.
Boardroom-Ready
Output is designed for senior leaders and boards — a score, a position, a roadmap. Not a 200-page deck.
Vendor-Independent
Calibrated to organisational reality, not retrofitted from any single vendor's product taxonomy.
Action-Oriented
Every assessment ends with a roadmap — what to do next, in what order, with what investment.
The five maturity levels
From ad-hoc to optimising.
The progression every AI-adopting organisation moves through. Progression is not linear — maturity must be earned at each level and defended over time.
Pilots happen by accident, not by design.
Some patterns emerging, no enterprise lens.
AI is now an organisational capability.
AI value is measured and managed.
AI is an enduring competitive advantage.
The five dimensions
AI maturity is not one number.
Five dimensions, each assessed separately. Almost every organisation has at least one dimension materially behind the others.
Where AI fits in business strategy and how it is sponsored.
Quality, accessibility, governance, and architecture of data.
AI platforms, infrastructure, tooling, and integration capability.
Skills, organisation design, ways of working, change management.
Risk, ethics, regulatory, controls, and accountability.
The AIMM Matrix
5 × 5 = 25 cells of evidence.
Every cell describes what maturity at this level on this dimension looks like concretely.
| Dimension ↓ / Level → | L1 — Initial | L2 — Developing | L3 — Defined | L4 — Managed | L5 — Optimising |
|---|---|---|---|---|---|
| Strategy | Reactive | Aligned | Documented | Measured | Strategic |
| Data | Silos | Catalogued | Integrated | Quality-managed | Data-as-product |
| Technology | Tools | Platform | Standardised | Operated | Innovating |
| People | Individuals | Pockets | Roles | Org design | AI-native |
| Governance | Ad-hoc | Emerging | Framework | Controls | Adaptive |
From score to roadmap
Insight without action is wasted.
Every AIMM assessment ends with a sequenced plan — the specific moves that take any single dimension from its current level to the next, prioritised against measurable criteria.
- 01Assess
Score each of the five dimensions against the rubric, anchored to observable evidence.
- 02Visualise
Plot the result on the AIMM radar. See where you lead, where you lag, where you are exposed.
- 03Sequence
Identify the lowest-cost, highest-leverage moves to lift the dimensions holding you back.
- 04Govern
Communicate to the board. Track progress quarterly. Defend maturity once earned.
Ready when you are
See where you sit — decide what comes next.
Run an AIMM assessment with your leadership team. Three hours. A scored position across five dimensions. A roadmap your board can act on.
Complementary to ISO/IEC 42001
The diagnostic behind the standard.
ISO/IEC 42001 specifies what a certifiable AI Management System must contain. AIMM tells you, honestly, how close your organisation is to satisfying it — and what to build next. The two are designed to be used together.
AIMM diagnoses
A scored, multi-dimensional view of where AI capability genuinely sits today — independent of any certification posture.
ISO 42001 certifies
An international management-system standard for responsible AI — auditable, certifiable, recognised by regulators and procurers.
Together they compound
AIMM sequences the work; ISO 42001 codifies it. Organisations using both reach certification faster and defend it more cheaply.
| AIMM Dimension | ISO/IEC 42001 Anchor | How they complement |
|---|---|---|
| Strategy | Clauses 4–6 — Context, Leadership, Planning | AIMM tells you whether your AI strategy is reactive or strategic; ISO 42001 requires you to document it as a managed system. |
| Data | Annex A.7 — Data for AI systems | AIMM grades how data is sourced, governed and integrated; ISO 42001 prescribes the controls that evidence it. |
| Technology | Annex A.6 — AI system life cycle | AIMM measures platform and life-cycle maturity; ISO 42001 codifies the life-cycle controls an auditor expects to see. |
| People | Clause 7 — Support, competence, awareness | AIMM diagnoses skills and operating model; ISO 42001 mandates the competence and awareness regime around them. |
| Governance | Clauses 8–10 + Annex A.2/A.3 — Operation, performance, impact | AIMM scores risk, ethics and accountability; ISO 42001 turns the score into an auditable management system. |
ISO/IEC 42001 is a registered standard of the International Organization for Standardization. AIMM is an independent, open framework and is not affiliated with ISO.
About AIMM
An open maturity model.
British by origin. Global by adoption.
Unlike maturity models authored behind a single corporate logo, AIMM is openly governed. It is co-developed and maintained as a public good — published under an open licence, shaped in the open, and free for any organisation, regulator or consultancy to apply.
UK Government
Co-developed with the Central Digital & Data Office, the AI Safety Institute, and the Government Digital Service as part of the national AI assurance agenda.
UK Public Agencies
Piloted across the NHS, HMRC, the Met Office, the Office for National Statistics and several local authorities to pressure-test the rubric at scale.
Industry Collaboration
Shaped with contributions from leading UK and global technology firms — cloud hyperscalers, model providers, systems integrators and domain specialists — all participating on equal footing.
Academic Partners
Research input from the Alan Turing Institute, Oxford, Cambridge, Imperial College London, UCL, Edinburgh and Manchester — keeping the framework grounded in independent evidence.
Origin
Forged in Whitehall. Refined in the open.
AIMM began as a working artefact inside UK central government, used to compare AI readiness across departments. Successive contributions from public bodies, universities and industry hardened it into a framework now used by organisations across Europe, North America, the Gulf, India, Australia and Singapore.
Governed openly
- Open specification — versioned, public, change-controlled.
- No vendor lock-in — any consultancy or in-house team may run it.
- Public revision cycle — contributions reviewed by an independent steering group.
- Aligned with ISO/IEC 42001, the EU AI Act and the UK AI assurance roadmap.
Open licence
Released under Creative Commons BY 4.0.
The AIMM specification, rubric and matrix are published under the Creative Commons Attribution 4.0 International licence. You may copy, adapt, embed and commercialise AIMM in your own assessments, products and publications — provided you attribute the AIMM framework and indicate any changes you have made.
Suggested attribution
"Based on the AIMM (Artificial Intelligence Maturity Model), © AIMM contributors, CC BY 4.0."
Originated in the United Kingdom. Maintained as a public good by an international community of public bodies, universities and industry contributors.