The enterprise AI readiness checklist: 12 questions before you build

Most AI pilots that disappoint were never ready to start. The model was rarely the problem. The use case was vague, the data was not accessible, or nobody owned the result. These twelve questions help you check readiness before you spend.

Value

1. Which business measure will this use case improve, and by how much would it need to move to be worth it?
2. Who are the users, and have they agreed the problem is worth solving?
3. What happens today, and how will we measure the before and after?

Data

1. Where does the data live, and can we access it legally and technically?
2. Is the data accurate and complete enough for this task?
3. Who owns the data, and who approves its use for AI?

Risk

1. Does the use case involve personal or confidential data, and which regulations apply?
2. What is the cost of a wrong answer, and where must a person review the output?
3. How will we log, audit and explain decisions?

People and cost

1. Who will own and run the solution after the pilot?
2. What skills do we need to build, buy or borrow?
3. What will it cost to run at scale, including model usage, and does the business case still hold?

If you can answer at least nine of these with confidence, you are ready for a focused pilot. If not, a short readiness assessment will close the gaps faster than a pilot that stalls.