📖 Content moderation glossary
What is human-in-the-loop moderation?
In short
Human-in-the-loop (HITL) describes an automated system in which a person reviews, confirms or corrects some of the machine's decisions, typically the uncertain, sensitive or contested ones.
In moderation, human-in-the-loop means the AI does not have the last word on everything. Cases below a confidence threshold, or in sensitive categories such as threats or self-harm, are routed to a person, who sees the content, its context and the reason the system flagged it.
Example: a comment says a politician should be shot down at the next election. A keyword-based system might read a threat; a person immediately understands the figure of speech and leaves it visible.
Human review also matters legally and reputationally. The Digital Services Act requires online platforms to handle complaints against moderation decisions under the supervision of appropriately qualified staff, not solely by automated means (Article 20). At OpenModeration, doubtful cases and contested decisions are reviewed by our team, and you see every decision in real time in one console.