The AI maturity model I actually use with clients

Published AI maturity models tend to have five or six named tiers, elaborate diagrams, and a vague sense that the goal is to move from the bottom tier to the top one over some unspecified timeline. They're pleasant to look at in a slide deck and rarely help a specific business decide what to actually do next.

The version I use with clients is three questions, not a tier diagram: is your data in a state where a system could actually use it without months of cleanup first, does your team have a documented process for the thing you want to augment or replace, and is there a specific person accountable for whether this initiative succeeds. Answering honestly places a business more usefully than any five-tier framework.

Most businesses that think they're behind on AI maturity are actually behind on the first question, data readiness, not on some absence of ambition or sophistication. And most businesses that think they're ready are underestimating how much the second and third questions, process documentation and clear ownership, actually determine outcomes more than the technology choice does.

A maturity model is only useful if it changes what someone does next. Three specific, answerable questions do that. A five-tier diagram, in my experience, mostly produces a satisfying sense of having done a strategic exercise without changing anyone's actual next step.