Stop measuring the rollout. Start measuring the capability.
88% of organisations are deploying AI. 86% of leaders say they were not ready for it. Gartner calls the gap the enablement illusion — access and adoption mistaken for transformation. These assessments measure what your workforce can actually do with AI, from readiness through to impact.
The rollout is ahead of the readiness.
Every major workforce study this year says the same thing: organisations moved faster on deploying AI than on knowing whether their people can use it.
Four stages of AI capability, measured org-wide.
Every organisation is somewhere on this path. Each stage answers a different question about the whole workforce — and answering them out of order tells you very little.
See where AI has actually landed across functions — and where licences were bought but practice never changed.
Pick the question you need answered.
Three measured across the organisation, ten measured per person. Select any one to see what it measures, when to use it, and what each side gets back.
AI Readiness
Whether the workforce can use AI safely and effectively before access widens.
The problemWe are giving everyone AI access, but we do not know who can use it safely.
Before widening AI access across teams — and whenever a new cohort, function or region is brought into the rollout.
A high score means people across the organisation can use AI productively without creating security, compliance or quality exposure.
CHRO · L&D · CISO and Risk · Head of AI · Transformation
Who gets AI access · Where to train first · Which teams carry exposure · Rollout sequencing
Shows exactly which teams sit below the safe-use bar, and where to train first, before AI is widened.
Shows each person their level per skill, what is strong, and the one module to close next.
See where your workforce actually stands.
Bring us one role or one team. We will show you the measurement, the evidence behind it, and what you would do differently on Monday.
Four steps, whichever test you pick.
Pick the decision first
Start from the decision you need to make — authorise AI use, hire, promote, redesign a workflow — not from the test name.
Assign to a role, not everyone
Each test names the functions it serves. Measure the population where the decision actually bites.
Read the level, not the number
Scores resolve to L1–L5 with a written meaning. A level is what you act on; a raw number is not.
Re-measure after development
A level is dated evidence. Close the gap, then re-measure to prove capability moved.
Measured once. Useful to both sides.
For the organisation
- Know who can be trusted to supervise AI before you authorise it
- Match oversight effort to risk instead of checking everything or nothing
- Stop plausible AI errors before they reach customers or regulators
- See genuine human contribution behind uniformly polished AI output
- Set automation boundaries deliberately, with accountability intact
- Defend AI-influenced decisions to auditors, boards and clients
For the individual
- Prove AI skills that a CV cannot evidence and AI cannot fake
- Get credit for original insight, not just polished output
- See exactly which level you are at, and what the next one requires
- Carry a dated, portable record of verified capability
- Target development at the specific construct that is weak
- Stand out for judgment and ownership, not tool familiarity
What buyers ask first.
How is this different from an AI literacy test?
Literacy tests ask whether someone knows what AI is and which tool to open. These measure what a person does with AI on real work — whether they verify it, when they override it, what they keep for themselves, and whether they can defend the result afterwards.
Can candidates just use AI to pass an AI assessment?
The tasks are built so that using AI is expected. What is scored is the judgment applied on top — which claim you checked, what you rejected, what you added, and whether you can defend the decision without the AI's words. Integrity checks and identity verification apply as they do to any GoMeasure assessment.
Do we need all ten?
No. Almost nobody does. Start from the decision in the chooser above; most organisations begin with one or two and add others as AI moves into more consequential work.
How do the levels work?
Every test resolves to the same five-level scale — Emerging, Developing, Proficient, Advanced, Strategic — with an interpretation written specifically for that construct. That makes results comparable across tests and across people.
Measure how your people actually work with AI.
Tell us the decision you need to make and we will tell you which test answers it.