Lesson
How to check an AI answer before using it.
A simple review routine for facts, context, tone, and risk.
Why this matters
Confidence comes from having a checking routine.
AI can sound polished when it is wrong. This gives the user a repeatable way to review an output before it becomes work product.
Use today
Ask “what would I need to verify?”
The habit is not trusting or rejecting AI blindly. The habit is knowing what needs checking.
The four checks.
Fact check
What might be wrong?
Look for claims, dates, numbers, or policy statements that need verification.
Context check
What is missing?
AI often answers confidently without knowing workplace context. Add context or edit manually.
Tone check
Does this sound right?
Make sure the output sounds like you and fits your workplace.
Risk check
Should this be used at all?
If the output affects people, money, legal, medical, HR, or confidential decisions, slow down.