Human-in-the-loop (HITL)

By Jorge García Aibar Updated 2026-09-25

The oversight mode in which a person can intervene in every decision cycle of an AI system ¹; in engineering terms, a gate that holds each consequential action until a named approver decides, logging approver, time to decide and override.

Developed in: ch. 05, Pattern: Human-in-the-loop Gate
ch. 11, From definition element to registry field
ch. 23, Human checkpoints and approval design
Chapters: ch. 05, Patterns · ch. 11, AI Defined · ch. 23, AI Agents
Contrast with: Human-on-the-loop (HOTL)
Source: 1 numbered reference, listed below

Commonly confused

Human-in-the-loop (HITL) versus Human-on-the-loop (HOTL)
Human-in-the-loop (HITL)
The oversight mode in which a person can intervene in every decision cycle of an AI system; in engineering terms, a gate that holds each consequential action until a named approver decides, logging…
Human-on-the-loop (HOTL)
The oversight mode in which a person can intervene in the design cycle and monitors the system's operation, rather than approving each decision.
The differenceA person approves each consequential decision, against a person monitors and can stop the system.
Why it mattersHITL is evidenced by approval logs and override rates; HOTL by alerting and a tested stop path.

Where it is used

13 chapters of the Body of Knowledge use the term. Each link opens the first section that does.

04 · The Stack Designing human oversight (Article 14) 2 mentions
05 · Patterns Pattern: Human-in-the-loop Gate 1 mention
08 · Regulatory Map EU AI Act, post-Omnibus 1 mention
11 · AI Defined From definition element to registry field 9 mentions
13 · Risk Management The loop: identify, assess, treat, monitor 2 mentions
14 · Development The use-case record 5 mentions
15 · Deployment Model types and deployment options 1 mention
16 · Fairness & XAI Monitoring fairness in production 2 mentions
17 · Incidents Root-cause analysis 1 mention
18 · EU AI Act High-risk requirements (Articles 8 to 15) 3 mentions
19 · Privacy & AI Automated decision-making 2 mentions
22 · Principles & Standards EU HLEG guidelines and ALTAI 5 mentions
23 · AI Agents Opening 3 mentions

Patterns that use this term

7 pattern pages use the term, most mentions first.

Human-in-the-loop Gate 2 mentions
Decision Notice & Contest Path 2 mentions
Runtime Guardrail 1 mention
Kill Switch / Circuit Breaker 1 mention
FRIA-as-Code 1 mention
Agent Identity & Scoped Credentials 1 mention
Explanation Artefact 1 mention

Sources

[1] Ethics Guidelines for Trustworthy AI (four ethical principles; seven requirements; human-in-the-loop, human-on-the-loop and human-in-command oversight). High-Level Expert Group on AI / European Commission. 2019-04-08. https://digital-strategy.ec.europa.eu/en/library/ethics-guidelines-trustworthy-ai (verified: primary)

Definitions of legal terms paraphrase the cited text, which governs. Dated statements are as of 2026-09-24.

Cite this term

García Aibar, J. (2026). Human-in-the-loop (HITL). In AI Governance Engineering: The Thesis & Body of Knowledge (v0.5.0), Glossary. https://doi.org/10.5281/zenodo.22956197. https://aigovernanceengineer.com/glossary/human-in-the-loop-hitl. CC BY 4.0

BibTeX

@misc{aige2026humanintheloophitl,
  author  = {Jorge García Aibar},
  title   = {{Human-in-the-loop (HITL)}},
  note    = {Glossary, AI Governance Engineering: The Thesis \& Body of Knowledge, version 0.5.0},
  year    = {2026},
  doi     = {10.5281/zenodo.22956197},
  url     = {https://aigovernanceengineer.com/glossary/human-in-the-loop-hitl}
}

Commonly confused