AArtificial Intelligence Maturity ModelVersion 2.1 · Open standard

Application · Assessment method

A defensible picture of current capability.

AIMM assessment is diagnostic rather than judgemental. It creates a shared factual baseline for governance, investment and improvement decisions.

Assessment cycle

The method combines documentary evidence, structured discussion and cross-functional moderation. A score without a traceable basis is treated as an opinion, not an AIMM result.

  1. 01Scope

    Define the organisational boundary, intended decisions, stakeholders and the period of evidence.

  2. 02Gather

    Collect policies, architecture, portfolio records, performance measures, interviews and observed practices.

  3. 03Test

    Compare evidence against each dimension-level description and identify contradictions or missing proof.

  4. 04Moderate

    Review provisional ratings across functions to separate isolated excellence from repeatable capability.

  5. 05Report

    Publish the maturity profile, confidence notes, material dependencies and evidence register.

Evidence register

Records the source, date, owner and relevance of material supporting each rating.

Confidence statement

Distinguishes well-evidenced positions from areas where further inquiry is required.

Moderation record

Preserves significant differences in interpretation and how the final rating was reached.