Sort each LLM risk scenario into the control family that best reduces it. Content boundary controls Data protection controls Human verification controls Tool permission controls A retrieved webpage tells the assistant to ignore system rules The model includes customer account IDs in a public answer A user forwards an AI legal summary without checking citations An AI agent can delete tickets without approval A hostile email tries to change the agent's instructions Training data contains secrets that can surface in output A fluent answer hides uncertainty about source quality A support bot can issue credits above policy limits
Read the full lesson
Sign up free: one personalized lesson every day, matched to your role and goals.
Already have an account? Sign in