human-in-the-loop
Human-in-the-loop is a design approach in which people review, approve or correct the work of an automated or AI system at set points.
Human-in-the-loop (HITL) describes systems that build human judgment into an otherwise automated process. In machine learning, it refers to people labeling data or rating model outputs so that the results feed back into training. In automation, it refers to a person checking and approving an important action before it runs.
In AI agent design, it serves as a safeguard that requires human approval before hard-to-reverse actions such as payments, data deletion or sending information outside an organization. A system that acts without human involvement is described as human-out-of-the-loop, and one in which people monitor and step in only when needed is described as human-on-the-loop.
This entry is based on AIPOST articles and widely known facts. If something is wrong, please send us a correction request.