GlossaryAgents and workflows

Responsive Salience

Responsive salience is when an AI interface automatically increases or decreases oversight UI—explanations, approvals, transparency—based on task risk and user context.

The system adapts how “loud” trust and control affordances are instead of one static autonomy setting.

What it means

High-risk or low-confidence tasks surface more gates and rationale; familiar low-risk tasks stay quiet so users are not nagged.

Why designers should care

Static UIs either over-warn (alert fatigue) or under-protect (incidents). Responsive salience matches oversight to moment-to-moment need.

Example

Deleting customer records triggers step-by-step confirmation and source quotes; summarizing internal notes runs with a slim progress indicator only.

Common mistakes

  • Same confirmation modal for low- and high-impact actions.
  • Changing salience with no user-visible reason.
  • No override when the system misjudges risk.

Explore more

Weekly AI UX notes

Patterns, prompts, and glossary updates for designers building AI products on Substack. No spam.

Subscribe on Substack