I’ve been noticing a pattern with how people use AI tools at work.
Not obvious misuse — just normal things like:
- debugging logs
- draft emails or proposals
- internal notes
- small pieces of client data
Individually it all feels harmless.
But when you step back, a lot of this is information that wouldn’t normally be sent to an external service.
It feels like there’s a gap between:
“don’t paste sensitive data”
and what actually happens day-to-day under time pressure.
Curious how others think about this.
Do you have any rules for what you will or won’t paste into AI?
[link] [comments]