The 10-Question AI Scorecard

Rules

Whether anyone would know if something went wrong.

1. Is there a written policy telling staff exactly what they can and can't put into an AI tool?

2. Could you say, right now, which of your data sets you legally can't afford to leak?

3. Do you know which AI tools — and which versions, free vs. paid — your people actually use?

4. Does anything get double-checked by a human before an AI-drafted fact goes out the door?

Whether AI has actually moved from a novelty to a habit.

5. Can every team lead name two repetitive tasks AI could take off their plate?

6. Have you moved past everyone experimenting alone, into a shared team or org-wide AI account?

7. Do you have any way of knowing whether the hours AI saves are landing somewhere, or just evaporating?

Whether AI is changing what you attempt, or just speeding up what you were already going to do anyway.

8. In the last month, has a leader on your team used AI to pressure-test a real decision — something riskier than drafting an email?

9. Have you talked, as a team, about how AI might change what you actually offer the people you serve?

10. Do your people have a place to try something with AI and fail, with nobody keeping score?