Design Workbook
The Psychology of AI User Experiences
Back to start
8/8
8. How to manage trust?
How will you help users gauge how much to trust the output? Select all that apply.
Capability reminders — Set expectations upfront about what the AI can and can't do. e.g., "I can sometimes be wrong — please double check."
Confidence levels — Signal how sure the AI is about a given output. e.g., Low / Medium / High.
Source transparency — Show where the output came from. e.g., "Based on data from [X, Y]."
Reasoning explanations — Tell the user why the AI did what it did. e.g., "I suggested this because…"
Human escalation — Offer a path to a real person. e.g., "Would you like to connect with a human expert?"
None — If not, explain why in the notes below.
Other — Describe in the notes below.
How will you handle errors?
Do you want to convey empathy, learning or something else?
NOTES: Explain why you made your choices.
Back
Next