Your question is Fix Confusion in an AI Feature. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
You shipped an LLM-powered feature and early user feedback says people are confused by its answers, confidence, or behavior. Some users are not sure what the feature can do, when to trust it, or why it responded a certain way.
How would you tackle a situation where user feedback indicates confusion with an AI feature?