Check your AI vocabulary

Nine questions about the vocabulary your firm actually uses. The reading comes from the 117 public definitions and the amendment record, not from a score of your firm. Nothing leaves this page. A blank submit, or an answer recorded as not sure, returns scope incomplete rather than a pass.

Do you keep a controlled AI vocabulary?Answer

Does your policy define AI system using a published definition?Answer

Is your AI inventory keyed to those terms?Answer

Is there a named owner for terminology?Answer

Do you map your terms to ISO/IEC 42001 clause 3?Answer

Are you tracking the generative AI amendment?Answer

Do you keep a change log for vocabulary changes?Answer

Do procurement and model providers use the same terms?Answer

When is your next vocabulary review?Date

The plan works back from the earlier of this date and the planning date, 16 February 2027.

What the reading is, and is not

It is a reading of nine answers against the public terms and the amendment record. It names a starting point, at least three actions and a dated plan. It is not a maturity score, not an audit and not a compliance result. The product does not certify, and no answer is sent anywhere.

Why some answers return scope incomplete

An unanswered question, or an answer recorded as not sure, returns scope incomplete and names the question. The engine does not default an unknown to the lighter pathway, because a reading built on an unknown is worse than no reading.

What happens to the answers

They stay in this browser tab. There is no account, no server call and no analytics event carrying an answer. The plan page keeps its rows in local storage; the check keeps nothing after the tab is closed.

Last updated: 2026-10-21 · Markdown version