Use maturity to locate a constraint, not award a badge.

Look across five dimensions

Commercial intent: priorities, value mechanisms and funding evidence.

Information and workflow foundations: source quality, access, integration and process clarity.

Governance and decision control: ownership, permissions, review and response.

People and adoption: skills, incentives, review capacity and practical use.

Delivery and operation: the ability to maintain, measure and change a deployed system.

These dimensions are a working lens used by Applied AI Australia, not an independently validated national index.

Expect uneven capability

An organisation may have strong engineering but weak benefit ownership. A team may use AI well while operating on poorly maintained information. One use case may be mature while another remains experimental.

A single averaged score can hide a critical gap. Keep the profile and evidence visible.

Test the basis of the assessment

Self-reported confidence differs from a demonstrated control or operating result. Ask what supports each answer and how current the evidence is.

A conversational assessment can prompt better evidence. It does not independently verify everything a participant supplies.

Avoid maturity as a compulsory ladder

More autonomy, custom models or a larger AI function are not automatically better. A constrained workflow using existing software may meet the objective.

Choose capability investment by the result it enables, the cost and the risk. Do not require every organisation to progress toward the same end state.

Turn the profile into a decision

Select the gap that blocks a valuable outcome. Name the owner, required evidence and next action. Where the gap is unowned or unfunded, the correct recommendation may be to pause the use case.