Generative AI at Work / Unit 2 GAI 048 worked example, 60 to 90 seconds
Which Risk Is This?
The idea
Naming a generative AI problem precisely is the first step to fixing it, and four questions turn a vague worry into one of NIST's twelve risks.
In practice
Step one: find the source. Did the trouble come from the model's output, from someone misusing the tool, or from how a person and the tool worked together? Step two: find the scope. Does it affect one use, the whole organization, or people outside it? Step three: match it to the list. A chatbot inventing a refund rule is confabulation; staff pasting customer records into a public tool is data privacy; a team rubber-stamping every draft is human–AI configuration. Step four: map it to the trustworthiness characteristics it threatens, which tells you what kind of fix to look for.
Where people slip
Many real problems carry two or more risks at once. Name each one; a single label hides the others.
Your place on the trail
Every lesson rests on earlier ones and opens later ones. Tap any to jump.
Something wrong in this lesson? Flag it · Sources for Generative AI at Work