Six levels. Six axes. One picture of where the practice actually sits, and the one move that would move it up.
One number cannot describe AI practice, because the practice is uneven by nature. A team strong on culture and weak on build capability is not the same team as one strong on build and weak on culture, even when a composite score puts them in the same bucket.
The model uses six independent dimensions on purpose. They are enough to describe the shape with fidelity and few enough to hold the picture in one head. The shape is the diagnosis.
Each axis runs from the practice you are starting from to the practice that defines the field.
Every team has a constraint axis: the dimension they have most neglected. The model's central operating claim is that the constraint sets the ceiling. You are only as AI-Ready as your weakest dimension.
This is the claim that does the work. It explains why tool-stuffed teams stall, why pilot programs proliferate without compounding, why “we use AI” can be true and irrelevant at the same time.
Three illustrative shapes. Same total area; different floors. The first runs into a wall; the second compounds; the third just looks impressive.
Six level names (Curious, Adopting, Building, Delegating, First, Native) exist to give a shared vocabulary to people who would otherwise argue about adjectives. The point of a level is not the level. The point is the move from one to the next.
The model names the plateaus where most teams pause. The next move is always the floor of the next step.
Models improve, tools change, the field moves. A score taken in March will not describe the same team in September. The level definitions hold. The bar inside each one rises with the field.
The model is built to be re-run on a 60-to-90-day cadence: frequent enough to catch movement, slow enough that the change is real.
They’re the ones with the highest floor.
Free. Ten minutes. A six-axis radar and three concrete next steps.