Human-in-the-loop (HITL)
A design in which a person has defined review, approval, intervention, or exception-handling responsibilities inside an AI-assisted process.
Why it matters
HITL should be explicit about when review is required and what evidence the reviewer receives.