AIPolicyTracker
ProcedureFree · no accountEU AI ActColorado ADMT law (SB 26-189)

Human Oversight Procedure

A procedure for the humans who oversee an AI system: what they must be able to do, when they intervene, how an override is recorded, drawn from every recorded oversight duty.

Formats: DOCX and XLSX · Version v1 · Built from dataset 914895c3103e · CC BY 4.0. You may use, adapt and share this template, including commercially, with attribution to aipolicytracker.org.

What's inside

  • Document: roles, competence, the oversight measures each recorded duty requires, intervention and override, escalation
  • Override log sheet: date, system, decision overridden, reason, reviewer

Preview

The sheets and sections of version v1, as built. Columns marked ▾ have a dropdown; ƒ is a formula.

Sheet: Override log · 9 columns · blank, 300 rows ready to fill
First rows of the Override log sheet
Log IDDateSystemOutput or decision overriddenAction taken ▾ReasonOverseerReviewed byReported as incident ▾
Rows are yours to fill; the dropdowns, formulas and colour rules are already in place.
Sheet: Oversight duties · 14 columns · 8 rows from the records
First rows of the Oversight duties sheet
DutyCategoryInstrumentJurisdictionWho it bindsNatureSource referenceApplies fromWhat it requiresEvidence a reviewer expectsISO/IEC 42001NIST AI RMFVerificationRecord
Test and monitor systems, enable human control, and be transparent with users (guardrails 4 to 6)Human oversightAustralian Voluntary AI Safety StandardAustraliaDeployer / user organisation, Provider / developerVoluntaryGuardrails 4, 5 and 6Test AI models and systems before deployment and monitor them in operation; enable meaningful human control and intervention; and inform end users about AI-enabTest reports and user disclosure recordsMEASURE and MANAGE functionsSource-linkedhttps://aipolicytracker.org/obligations/australia-vaiss-testing-human-control-transparency
Enable and assign effective human oversightHuman oversightEU AI ActEuropean UnionProvider / developer, Deployer / user organisationLegal requirementArticle 14; Article 26(2) for deployers2026-08-02High-risk systems must be designed with human-machine interface tools so natural persons can effectively oversee them, understand capacities and limitations, avHuman oversight procedure and role assignmentAnnex A control on human oversightGOVERN 3.2, MANAGE 2.xSource-linkedhttps://aipolicytracker.org/obligations/eu-ai-act-human-oversight
Employers and employment agencies must let candidates request an alternative selection process or accommodationHuman oversightNYC Local Law 144 (automated employment decision tools)New York (United States)Deployer / user organisationLegal requirementNYC Administrative Code Section 20-871(b)(1); 6 RCNY Section 5-3032023-07-05The advance notice to candidates and employees must allow them to request an alternative selection process or a reasonable accommodation. The DCWP rules make clAlternative-process and accommodation request logAnnex A.9.2GOVERN 5.1, MANAGE 4.1Verified against the official source 26 Sep 2026https://aipolicytracker.org/obligations/us-new-york-city-local-law-144-alternative-process-request
Determine the appropriate level of human involvement in AI decisionsHuman oversightSingapore Model AI Governance FrameworkSingaporeDeployer / user organisationVoluntarySecond edition, Part on human involvement in AI-augmented decision-makingUsing a risk-impact matrix (probability and severity of harm), organisations choose human-in-the-loop, human-over-the-loop or human-out-of-the-loop designs and Risk-impact assessment and human-involvement decisionGOVERN 3.2, MANAGE 2.xSource-linkedhttps://aipolicytracker.org/obligations/singapore-mgf-human-involvement
Operators of high-impact AI must ensure human management and supervisionHuman oversightFramework Act on the Development of Artificial Intelligence and Establishment of a Foundation for TrustSouth KoreaProvider / developer, Deployer / user organisation, Public authority / government bodyLegal requirementArticle 34(1)2026-01-22Operators of high-impact AI must ensure that the AI is subject to human management and supervision, so that people remain able to monitor its operation and inteHuman supervision design and assigned supervisorsAnnex A.9.2GOVERN 3.2, MANAGE 2.4Verified against the official source 26 Sep 2026https://aipolicytracker.org/obligations/south-korea-ai-basic-act-art-34-high-impact-ai-human-oversight
Respect the right to object to automated decision-making without human interventionHuman oversightUAE PDPLUnited Arab EmiratesDeployer / user organisationLegal requirementArticle on data-subject rights relating to automated processing (reviewer to cite article number)Data subjects may object to decisions based solely on automated processing, including profiling, that produce legal or similarly serious effects, subject to excAutomated decision objection procedureGOVERN 5.x, MANAGE 4.xSource-linkedhttps://aipolicytracker.org/obligations/uae-pdpl-automated-decision-objection

Document outline (DOCX)

  1. 1. Purpose
  2. 2. Roles and competence
  3. 3. What the overseer must be able to do
  4. 4. When to intervene
  5. 5. Recording an intervention
  6. 6. Duties this procedure serves
  7. Test and monitor systems, enable human control, and be transparent with users (guardrails 4 to 6)
  8. Enable and assign effective human oversight
  9. Employers and employment agencies must let candidates request an alternative selection process or accommodation
  10. Determine the appropriate level of human involvement in AI decisions
  11. Operators of high-impact AI must ensure human management and supervision
  12. Respect the right to object to automated decision-making without human intervention
  13. Provide routes to contest AI outcomes and seek redress
  14. Apply safeguards to solely automated decisions with significant effects

Duties this template covers (9)

Each is cited in the file with its source reference and a link back to the record.

Legal basis

Version history

Versions of Human Oversight Procedure
VersionBuiltDatasetWhat changed
v1914895c3103eFirst version, built from dataset 914895c3103e.

Only the latest version is served. A rebuild that changes the content adds a version; a rebuild that does not is skipped.

Frequently asked questions

Is the Human Oversight Procedure free?
Yes. Download the DOCX and XLSX without an account, under CC BY 4.0. You may use, adapt and share this template, including commercially, with attribution to aipolicytracker.org.
What is it generated from?
Version v1 was built on 26 September 2026 from dataset 914895c3103e: 16 recorded duties are cited in it, drawn from 9 instruments. Every row that cites a duty links to the record, and the record links to the official source.
How will I know when it changes?
The library is rebuilt daily. When a change to the records reaches this template it gets the next version, a changelog in the version history below, an entry in the AI policy updates hub and the templates feed, and a line in the weekly digest for subscribers of the templates topic.
Does completing it make us compliant?
No. It is an informational resource, not legal advice; it helps produce the evidence a regulator, customer or auditor asks for. Whether a duty applies to you is a judgement the template cannot make.

Informational only, not legal advice. Verify every claim against the linked official sources and consult qualified counsel before acting.

Frequently asked questions

Is the Human Oversight Procedure free?
Yes. Download the DOCX and XLSX without an account, under CC BY 4.0. You may use, adapt and share this template, including commercially, with attribution to aipolicytracker.org.
What is it generated from?
Version v1 was built on 26 September 2026 from dataset 914895c3103e: 16 recorded duties are cited in it, drawn from 9 instruments. Every row that cites a duty links to the record, and the record links to the official source.
How will I know when it changes?
The library is rebuilt daily. When a change to the records reaches this template it gets the next version, a changelog in the version history below, an entry in the AI policy updates hub and the templates feed, and a line in the weekly digest for subscribers of the templates topic.
Does completing it make us compliant?
No. It is an informational resource, not legal advice; it helps produce the evidence a regulator, customer or auditor asks for. Whether a duty applies to you is a judgement the template cannot make.