Your leaders are confident about AI. Are they right?
Twenty minutes each, across six domains that decide whether an AI programme survives contact with reality. You get a team score, the spread behind it, and the distance between how capable your leaders feel and how capable their answers say they are.
The cheapest way to find out what your leadership team does not know before it costs you something. Free to run.
Real screens · sample team
The problem
Nobody in a leadership meeting says they do not understand AI.
So the strategy gets agreed by a room where several people are quietly guessing, and nobody finds out which ones until a vendor oversells something, or a customer asks a question that should have had an answer.
Confidence is not capability
The two get asked as one question in most organisations, which means the answer is always confidence. Ask them separately and the distance between them is often twenty points.
An average hides the problem
A leadership team averaging Developing might be uniformly middling, or it might be one strong leader carrying two who are nowhere. Those need completely different responses.
Governance is where it falls over
Not the technology. Who may adopt a tool, who signs off before AI output reaches a customer. Teams routinely discover they each believed someone else owned it.
How it runs
You set it up once. It chases everybody else.
You add your leadership team
Name and email for each person. Eight to fifteen is typical: the executive team, or the layer below it, or both if you want to see where the two diverge.
Everyone gets their own link
A personal invitation in your name, with a link that takes them straight into their assessment. Reminders go out on their own; you can see who has finished without chasing anybody.
You get the team picture, not just the average
A team score with the spread behind it, the gap between how confident they feel and what they can actually do, and the domains where your leaders quietly disagree with each other.
Capability statements
Scored one to five on how true each is of them today. "I know who is accountable for signing off before AI-generated output reaches a customer." Specific enough that guessing is hard.
A confidence check
Asked separately from capability, and deliberately so. Comparing the two is what turns a score into a finding.
Evidence questions
Short written answers, read rather than scored. What AI question are they most often asked that they cannot answer well? These are usually the most quoted part of the report.
What you get
Three findings, and the third is the one people remember.
These are the actual screens, filled with a sample team rather than a customer's.
Where the team sits, and how far apart they are
A score out of 100 in one of five bands, and next to it the distribution. An average of 57 means something quite different when one leader is Emerging and another is Strong, and the distribution is what tells you whether you have a training problem or a hiring one.
The six domains, and where your leaders disagree
One dot per leader on each domain. Where the dots cluster, the team has a shared view, right or wrong. Where they scatter, your leaders believe different things about how this organisation works, which is worth knowing before you ask them to agree on a plan.
- AI fundamentals and literacy — Whether they could spot an AI proposal that is technically impressive and commercially pointless.
- Value identification and business case — Whether they understand what running an AI system costs, not just what building one costs.
- Data, privacy and risk — Whether they could tell a customer today how AI touches that customer's information.
- Governance and decision rights — Whether they know who decides that a new AI tool may be adopted, and who signs off before AI output reaches a customer.
- Delivery, vendors and build versus buy — Whether they can test a vendor's AI claims, and judge whether to build, buy or wait.
- Leading change and building capability — Whether they have actually told their team what they may and may not use AI for.
The finding
The gap between feeling ready and being ready.
Confidence and capability are asked as separate questions, which almost nothing else does. When a team rates its confidence well above what its answers support, it is not a character flaw. It means decisions are being made with more certainty than the underlying knowledge justifies, and nobody in the room can tell.
A calibrated team is not necessarily a capable one. But a team that knows where it is weak asks better questions of its vendors, its CTO and itself — and that is the thing you can fix in a quarter.
Find your team's gapWho it is for
Anyone about to bet on a team's AI judgement.
CEOs and executive teams
Before you commit to an AI strategy, a read on whether the people who have to lead it could defend it. The gap between confidence and capability is usually the surprise.
Boards
Evidence for the oversight question you are meant to be asking: does management actually know who governs AI here, and would they answer the same way as each other?
CTOs and CIOs
A way to show, rather than assert, that the rest of the leadership team needs to come with you. And a list of exactly which topics to spend the training budget on.
What it costs
Run your first round for nothing
Create an account, add your leaders, and send the invitations. You get the whole report: team score, distribution, the six domains, the divergence, the confidence gap and every written answer.
- No card, no sales call, no trial clock
- Invitations and reminders sent for you
- Export the report as a PDF or an executive deck
- Re-run it after training to show what moved
Want the findings turned into a plan? The AI Leadership Sprint is four weeks with an independent CTO, and usually starts here.
Questions
The things people ask first
Ask the six questions nobody asks out loud.
Twenty minutes each. You will know within a week whether your leadership team is as ready as it thinks it is.