World Model Readiness
Engraved role-check instrument

For you

Role Checks

Module · Can you question what you approve

The Board Member's AI Check

You are asked to approve AI strategies, budgets and risk appetites, and your signature carries a duty of oversight that does not pause for a technology you find unfamiliar. The danger is not that you lack a data science degree; it is nodding along to a management narrative you cannot test. This module checks the five things that let you govern rather than rubber-stamp: your own literacy, the questions you can ask, independence from the story you are told, a clear view of risk appetite, and a sense of what good looks like elsewhere.

Question 1 of 5 · You understand enough

Do you personally understand AI well enough to govern it, not just approve it?

You do not need to build models, but you do need to know what they can and cannot do, where they fail, and what the words in the board pack actually mean. Oversight you cannot exercise is oversight in name only.

Question 2 of 5 · You have the right questions

When management presents an AI plan, do you know what to ask?

The most valuable thing a director brings is the question that exposes an unexamined assumption. What is the failure mode, who is accountable, what does it cost when it is wrong, how do we know it works: if you cannot ask these, you cannot test the plan.

Question 3 of 5 · You think independently

Can you form a view on an AI initiative independent of what management tells you?

Management writes the board pack, chooses the metrics and frames the story. That is not a criticism, it is the structure, and it is exactly why a director needs an independent read. A board that only knows what it is shown is governed by the people it is meant to govern.

Question 4 of 5 · You can name the risk appetite

Could you state, in plain terms, how much AI risk this company should accept?

Setting risk appetite is a board duty, not a management one. If you cannot say where the line sits, how much autonomy is too much, what failure is unacceptable, then the line is being drawn operationally by whoever ships next, without you.

Question 5 of 5 · You know what good looks like

Do you have any outside reference for whether this company's AI approach is sound?

Judged only against its own plan, every initiative looks reasonable. Calibration comes from outside: peer boards, other sectors, how comparable organisations handle the same risk. Without it you cannot tell ambition from recklessness or caution from paralysis.

For the statistics · one click each

Three questions for the public picture

These do not affect your score. They feed the anonymised, aggregated statistics; groups under 8 respondents are never shown.

How often does AI reach your board's agenda as a real item?

Never
Only after incidents
Occasionally
A standing item
Central to strategy

Where does your own AI understanding come from?

Nowhere in particular
Mostly management
General media
Deliberate education
My own background

Can the board get AI advice independent of management?

No
Informal contacts
Occasionally commissioned
A standing arrangement
I am not sure

Your context

Used to calibrate the report. Company size and sector remain in the anonymized dataset; your email does not.

What the five levels look like

Every dimension in this assessment is scored 1 to 5. This is what the levels mean, dimension by dimension. The graded report diagnoses where your own answers land and what to do about it.

You understand enough

  1. 1Lost in the jargon
  2. 2Follow the headlines
  3. 3Grasp the basics
  4. 4Confident on the concepts
  5. 5Fluent enough to probe

At the low end: If the board pack's language loses you, you cannot discharge the oversight duty your seat carries. Invest a few hours in the fundamentals; a director who cannot follow the risk cannot govern it. What good looks like: Being fluent enough to probe is exactly the bar for a director. Keep it current; the capability and the risks both move faster than the annual training cycle.

You have the right questions

  1. 1No idea what to ask
  2. 2Generic questions
  3. 3A few good ones
  4. 4A reliable repertoire
  5. 5Questions that change plans

At the low end: A board that cannot question an AI plan simply ratifies it. Build a short list of standing questions on failure, accountability and evidence, and ask them every time. What good looks like: Questions that actually change plans are the mark of a board doing its job. Keep sharpening the repertoire; the useful question is often the one management hoped you would not ask.

You think independently

  1. 1Take the pack as given
  2. 2Rarely challenge it
  3. 3Question the framing
  4. 4Seek my own inputs
  5. 5Independent, evidenced view

At the low end: Taking the board pack as the whole truth hands your oversight to the people you oversee. Ask for the data behind the summary, and for the risks the narrative left out. What good looks like: An independent, evidenced view is what makes your approval mean something. Protect the habit; the more confident the narrative, the more it deserves an outside check.

You can name the risk appetite

  1. 1Never discussed
  2. 2Vague sense
  3. 3Discussed, not decided
  4. 4Agreed in principle
  5. 5Articulated and documented

At the low end: An unstated risk appetite gets set by default, one deployment at a time, far below your seat. Put it on the agenda: what would we never let AI decide, and how much are we willing to lose to move fast. What good looks like: An articulated, documented risk appetite is what lets the company move fast safely, because everyone knows where the edge is. Revisit it as the technology and the stakes change.

You know what good looks like

  1. 1No outside reference
  2. 2Anecdotes only
  3. 3Some peer awareness
  4. 4Deliberate comparison
  5. 5Actively calibrated

At the low end: With no external reference, you are grading the company against its own homework. Seek out how peer boards handle AI risk; the contrast is where your judgment sharpens. What good looks like: Actively calibrating against the outside world is what lets you tell bold from reckless. Keep the aperture open; the reference points that mattered last year may already have moved.