Generative AI at Work / Unit 4 GAI 084 short lesson, about 20 seconds
No Secrets in Prompts
Captions available
The idea
The DAIR.AI guide documents prompt leaking: attacks that coax a model into revealing its hidden instructions, including anything confidential written into them.
In practice
A shared team assistant whose system message includes an internal discount code, which a determined user extracts.
Where people slip
Write every prompt, including system messages, as if its users could read it.
Your place on the trail
Every lesson rests on earlier ones and opens later ones. Tap any to jump.
Unlocks
- Nothing yet depends on this.
Something wrong in this lesson? Flag it · Sources for Generative AI at Work