Decides what a business should and should not build with AI, and in what order.
What separates them
Anyone can talk about strategy; the ones worth hiring have actually killed a funded project and can explain exactly why.
Ask these
01
Tell me about a project you recommended stopping after it had already been funded.
Tests whether they have real authority and will use it against sunk cost, rather than just advising in the abstract.
A strong answer
Names a specific project, the financial or political cost of stopping it, the evidence that triggered the decision, and what happened afterward.
The confident failure
Talks generally about being data-driven and rigorous in evaluation, but the project they eventually name was never actually funded or was easy to cancel.
02
How do you decide what a business should not build with AI?
Tests whether they have real filtering criteria or are just following whatever is fashionable.
A strong answer
Gives concrete criteria such as data readiness or the quality of the existing process, and describes a real case where they recommended a simpler fix instead of an AI approach.
The confident failure
Talks about starting small and proving value first, without naming any concrete criterion that would actually rule something out.
03
How would you decide which of several competing AI proposals to fund first?
Tests prioritisation logic rather than a general enthusiasm for AI initiatives.
A strong answer
Names concrete criteria such as data availability, the cost of getting it wrong, and who owns the outcome, and gives an example of a real prioritisation call.
The confident failure
Talks vaguely about aligning proposals with strategic pillars without describing how a real decision was actually made.
04
What is your process for checking, after launch, whether an AI initiative delivered what was promised?
Tests whether they follow through after the pitch rather than moving on to the next project.
A strong answer
Describes a metric agreed before launch, a review cadence, and gives a real example of an initiative being judged a failure and shut down.
The confident failure
Says they track return on investment closely, without being able to name what was measured or when it was reviewed.
What we ask when assessing for the register
Harder, and answerable only by somebody who has done the work. Published because a question that stops working when it is known was never testing anything.
01
Describe a case where the data or technology existed to build something, but you advised against it anyway. What was the real objection?
Distinguishes genuine business judgement from technology-first thinking dressed up as caution.
A strong answer
Gives a genuine business reason such as the organisation not being ready to act on the output, or the cost of a wrong decision being too high, and can describe the decision trail that led there.
The confident failure
Gives a vague answer about risk without naming a real business mechanism or a decision that anyone else would remember.
02
How do you tell the difference between an AI problem and a broken process dressed up as one?
Many requests for AI are really requests to fix something that was never designed properly.
A strong answer
Describes specific diagnostic questions they ask, and gives an example of redirecting a request from an AI build to a process fix instead.
The confident failure
Says they always look at the data first, but the answer conflates data quality with process design and never resolves into a real case.
Ask these whatever the discipline
Tell me about something you built that failed in production. What broke, how did you find out, and what did you change?
What would you refuse to do on this project, and what would you tell me instead?
How would you know, three months in, that this was not working?
What is the part of your own work that you are least confident about?
Relevant here, and none of them is evidence on its own. What each does and does not prove is set out in full on the certifications page.
Foundation Certificate in Artificial Intelligence, BCS, The Chartered Institute for IT. Anything at all about building or assessing AI systems. BCS states there are no entry requirements, and the certificate is honest about being a foundation.