HUMAN–AI UX
Design the boundary between what AI can do and what people should control.
Human–AI UX is not only interface copy. It is the operating model for intent, permission, autonomy, evidence, intervention and recovery.
A PRACTICAL MODEL
Consequence and reversibility determine control.
Low-risk, reversible actions can be automated with light supervision. High-impact, irreversible or novel actions need stronger previews, approval and recovery. The interface should make those boundaries legible.
- Scope before the AI starts.
- Progress while tools and data are being used.
- Evidence when results affect decisions.
- Correction and undo when output is wrong.
- Human handoff when automation should stop.