← AI Glossary

Human-in-the-loop (human oversight)

Ethics & compliance

Human-in-the-loop describes an AI system in which a person steps in at key stages to validate, correct or take back control before a sensitive decision is applied. The AI Act requires effective human oversight for high-risk AI systems.

In practice at Gensai

AI agents built by Gensai include guardrails and human oversight for any action that commits the business.