On audit trails and user trust

Something I keep noticing in user research sessions: when you show people why the AI made a recommendation, even a simple log of inputs and reasoning, their trust jumps noticeably.

This is counterintuitive. You would expect that exposing the machinery makes people more skeptical. Instead, transparency creates a sense of control.

The key insight here is that audit trails are not just for regulators. Surface them in the UI. A simple "View decision log" link does more for user confidence than any amount of accuracy marketing.