Information boundary check
Can this information be used with AI?
Describe the category, not the actual information. This quick check gives a practical boundary and a safer alternative before anyone pastes sensitive work into an AI tool.
Boundary
Use an approved low-risk example
Do this instead
Human check
Why the context matters
One tool is not one risk level.
Information, access controls, intended use, and human review all change the risk. When in doubt, pause rather than trying to make a sensitive record fit a public AI tool.