← Full glossary
Human-in-the-loop (HITL)
PatronHITL · human oversight
Human-in-the-loop (HITL) is the pattern where a person approves, corrects or supervises an agent's decisions before or during the action, instead of letting it act alone. It is the practical lever for giving an agent autonomy without giving up control.
HITL is not all-or-nothing: it is a dial. You can require human approval only above a threshold (amount, risk, low model confidence), review a sample of actions, or have the human step in only on exceptions. Done well, it lets you start with heavy oversight and loosen the reins as the agent proves reliability with data. It is the mechanism that makes autonomy gradual and auditable, not a leap of faith.
How it differs from
- Full autonomy
- In full autonomy no person is in the decision; HITL keeps a human control point, fixed or by exception.
Examples
- Having an agent draft the reply but a human approve it before it reaches the customer is HITL.
- Asking for human confirmation only when the agent's confidence drops below a threshold.
FAQ
- Does HITL kill the agent's efficiency?
- Only if you apply it to everything. The point is to place it where risk justifies it (high amounts, rare cases) and let the agent run on its own for routine, well-measured work.