.png&w=1536&q=75)
Free Lesson
How To Design For Trust and Confidence In AI Products
90 min
Apr 2, 2026 12:00 PM
Virtual (Zoom)
In this video
What you'll learn
How to calibrate trust in AI to avoid costly mistakes
Why AI products shouldn't aim for maximum trust, but just enough trust to be reliable.
How to handle AI hallucinations and wrong assumptions
We can't fully prevent them, but we can set up proper guardrails, human-in-the loop, user's control, verification.
Why it's a good idea to avoid confidence scores
Useful design patterns such sandboxing, intent confirmation, consensus meter, context engineering.
Why this topic matters
We often assume that just because a product has AI features, users will automatically use it for their daily work. People expect software to work reliably — but it's just a matter of time until AI quality starts drifting. In this session, we'll study how to design, build, calibrate and maintain user's trust, repeatedly, over time.
You'll learn from
Vitaly Friedman
20 years of experience • Senior UX Lead @ EU Parliament • Speaker • SmashingMag
Previously at
