Skip to main content

AIGO — AI Governance Operating Framework

AI Governance Template

Version: 0.2 Status: Draft Working Name: AIGO Full Name: AI Governance Operating Framework Document Identifier: AIGO-TPL-001 Document Type: Governance Template Template Purpose: Organizational AI Governance Arrangement

1. Template Purpose

This template provides a structured record for establishing, documenting, approving, operating, reviewing, and improving an organization’s AI governance arrangement under the AIGO AI Governance Operating Framework. The completed record should establish:
  • governance scope;
  • governance objectives;
  • governance principles;
  • governance authority;
  • roles and responsibilities;
  • decision rights;
  • AI portfolio governance;
  • risk governance;
  • lifecycle governance;
  • control governance;
  • monitoring;
  • assurance;
  • incident governance;
  • change governance;
  • risk acceptance;
  • retirement;
  • evidence and records;
  • reporting;
  • regulatory and external requirements;
  • third-party governance;
  • exceptions;
  • continual improvement.
This template is intended to be completed using the organization’s actual governance arrangements. It is not itself a governance policy and does not replace the applicable AIGO procedures.

2. Template Completion Instructions

Complete the fields that apply to the organization’s governance arrangement. Where a section is not applicable, record: Not Applicable — [reason] Where a decision, requirement, control, or activity is delegated to another controlled record, reference that record rather than duplicating it. Use controlled identifiers wherever possible. Recommended identifiers include:
  • Governance Record ID;
  • AI System ID;
  • Risk ID;
  • Control ID;
  • Approval ID;
  • Incident ID;
  • Change ID;
  • Evidence ID;
  • Assurance ID;
  • Improvement ID.

3. Governance Record

3.1 Governance Record Identification

Governance Record ID: Organization: Business Unit / Function: Governance Arrangement Name: Governance Owner: Prepared By: Date Created: Last Updated: Version: Status:

3.2 Governance Arrangement Status

Current Status:
  • Draft
  • Under Review
  • Approved
  • Approved with Conditions
  • Suspended
  • Under Revision
  • Retired
Effective Date: Next Review Date: Approval Authority:

4. Governance Scope

4.1 Organizational Scope

Organization covered by this governance arrangement:

4.2 Business Scope

Business functions, departments, products, or services covered:

4.3 Geographic Scope

Countries, regions, jurisdictions, or legal entities covered:

4.4 AI Scope

AI systems, AI-enabled processes, AI services, AI development activities, and AI-related third parties covered:

4.5 Lifecycle Scope

AI lifecycle activities covered:
  • Govern
  • Identify
  • Classify
  • Assess
  • Treat
  • Approve
  • Deploy
  • Operate
  • Monitor
  • Assure
  • Improve
  • Change
  • Continue
  • Retire
Applicable lifecycle scope:

4.6 Exclusions

Activities, systems, business functions, jurisdictions, or other areas excluded from scope: Exclusion rationale:

4.7 Scope Review

Scope review frequency: Scope-change trigger:

5. Governance Context

5.1 Organizational Context

Relevant organizational objectives: Organizational context affecting AI governance:

5.2 AI Governance Context

Why AI governance is required within the organization: Key AI-related opportunities: Key AI-related risks:

5.3 Stakeholder Context

5.4 External Context

Relevant external factors may include:
  • laws;
  • regulations;
  • standards;
  • contractual obligations;
  • industry requirements;
  • customer expectations;
  • technology dependencies;
  • third-party dependencies.
External context:

6. Governance Objectives

6.1 Primary Objectives

The governance arrangement should define measurable or assessable objectives. Primary objectives:

6.2 Expected Governance Outcomes

Expected outcomes:

6.3 Governance Success Criteria

How governance effectiveness will be determined:

7. Governance Principles

The governance arrangement should identify the principles that guide AI-related decisions.

7.1 Applicable AIGO Principles

The following principles may be adopted or adapted:
  • Accountability
  • Transparency
  • Human Oversight
  • Risk-Based Governance
  • Proportionality
  • Security
  • Privacy
  • Fairness
  • Safety
  • Reliability
  • Resilience
  • Traceability
  • Evidence-Based Decision-Making
  • Lifecycle Accountability
  • Continual Improvement
Additional principles:

7.2 Principle Application

How the principles are applied in practice:

8. Governance Authority

8.1 Primary Governance Authority

Governance Body / Authority: Mandate: Authority: Accountability:

8.2 Executive Sponsorship

Executive Sponsor: Executive Responsibilities:

8.3 Delegated Governance Authority

Delegated authority: Delegation scope: Delegation conditions: Delegation limitations: Delegation expiry / review:

9. Governance Structure

9.1 Governance Bodies

9.2 Reporting Structure

Governance reporting relationships: Escalation relationships:

9.3 Governance Committee Requirements

Quorum requirements: Decision-record requirements: Conflict-of-interest requirements: Meeting-record requirements:

10. Governance Roles and Responsibilities

10.1 AI Governance Owner

Role / Position: Responsibilities:
Authority: Escalation Path:

10.2 AI System Owner

Role / Position: Responsibilities:
Authority: Escalation Path:

10.3 Risk Owner

Role / Position: Responsibilities:
Authority: Escalation Path:

10.4 Control Owner

Role / Position: Responsibilities:
Authority: Escalation Path:

10.5 Assurance Role

Role / Function: Responsibilities:
Independence Requirements:

10.6 Other Roles


11. Segregation of Duties

The governance arrangement should identify activities requiring separation of responsibilities.

11.1 Segregated Activities

11.2 Conflict-of-Interest Requirements

Requirements:

12. Decision Authority

12.1 Governance Decision Matrix

12.2 Delegated Decisions

Delegated decision authorities: Conditions and limitations:

13. AI Portfolio Governance

13.1 AI Inventory

AI Inventory Owner: Inventory Location / System: Inventory Review Frequency: Inventory Reconciliation Method:

13.2 Inventory Minimum Information

Each AI system should be identifiable through, as applicable:
  • AI System ID;
  • System Name;
  • System Owner;
  • Business Owner;
  • Intended Purpose;
  • Classification;
  • Risk Level;
  • Lifecycle Stage;
  • Deployment Status;
  • Applicable Controls;
  • Approval Status;
  • Monitoring Status.

13.3 Portfolio Review

Portfolio review frequency: Portfolio risk review method: Aggregate-risk review requirements:

14. AI System Governance

14.1 Registration Requirement

Each AI system within scope should have a controlled registration record. Registration Procedure: Registration Owner: Registration Repository:

14.2 Required Registration Information

  • AI System ID
  • System Name
  • System Owner
  • Business Owner
  • Intended Purpose
  • AI Capability
  • Classification
  • Risk Level
  • Lifecycle Stage
  • Deployment Environment
  • Data Categories
  • Model / Provider
  • Third-Party Dependencies
  • Applicable Controls
  • Approval Status
  • Monitoring Status

14.3 System Profile

System Profile Requirement: System Profile Repository: Profile Review Frequency:

15. AI Classification Governance

15.1 Classification Method

Classification Methodology: Classification Owner: Classification Approval Authority:

15.2 Classification Criteria

Classification may consider:
  • intended purpose;
  • affected stakeholders;
  • individual impact;
  • business criticality;
  • autonomy;
  • decision significance;
  • potential harm;
  • data sensitivity;
  • security;
  • privacy;
  • fairness;
  • safety;
  • regulatory requirements;
  • deployment scale;
  • reversibility;
  • human oversight.

15.3 Classification Record

AI System ID: Classification: Classification Date: Classification Owner: Reviewer: Approval Authority: Supporting Evidence IDs: Classification Rationale:

15.4 Reclassification Triggers

Reclassification should be considered when:
  • intended purpose changes;
  • decision significance changes;
  • automation changes;
  • affected populations change;
  • material data changes occur;
  • material model changes occur;
  • risk changes materially;
  • significant incidents occur;
  • applicable requirements change.
Additional triggers:

16. AI Risk Governance

16.1 Risk Management Method

Risk Methodology: Risk Scoring Method: Risk Criteria:

16.2 Risk Appetite

AI Risk Appetite:

16.3 Risk Tolerance

AI Risk Tolerance:

16.4 Risk Ownership

Risk Owner Assignment Method:

16.5 Risk Escalation

Risks should be escalated when:
  • risk exceeds tolerance;
  • required controls are unavailable;
  • critical controls fail;
  • material incidents occur;
  • significant changes occur;
  • monitoring thresholds are exceeded;
  • uncertainty becomes material;
  • legal or regulatory concerns arise.
Escalation Process: Escalation Authority:

17. AI Risk Acceptance Governance

17.1 Acceptance Requirements

Residual risk may be accepted only when:
  • the risk has been assessed;
  • treatment options have been considered;
  • applicable controls have been evaluated;
  • residual risk is understood;
  • acceptance is within authorized tolerance;
  • appropriate authority approves the decision;
  • the decision is documented;
  • review requirements are established.

17.2 Acceptance Authority

Risk Acceptance Authority:

17.3 Acceptance Conditions

Conditions for acceptance:

17.4 Acceptance Monitoring

Monitoring Requirements: Review Frequency: Expiry / Reapproval Requirements:

18. AI Lifecycle Governance

Governance should operate throughout the AIGO lifecycle.

18.1 Lifecycle Governance Requirements

Requirements:

18.2 Lifecycle Entry and Exit Criteria


19. Governance Controls

19.1 Control Framework

Applicable Control Framework: Control Repository:

19.2 Control Ownership

19.3 Control Exceptions

Exception Process: Exception Approval Authority: Compensating Control Requirements: Exception Monitoring:

20. Human Oversight Governance

20.1 Oversight Requirements

Required Human Oversight:

20.2 Oversight Responsibilities

Responsible Role: Oversight Activities:

20.3 Human Authority

Can human users challenge AI outputs? Can human users override AI outputs? Who has final decision authority?

20.4 Escalation

Conditions requiring human escalation:

21. AI Monitoring Governance

21.1 Monitoring Requirements

Monitoring Scope: Monitoring Owner: Monitoring Frequency: Enhanced Monitoring Triggers:

21.2 Monitoring Indicators

21.3 Monitoring Evidence

Evidence Repository: Evidence Retention:

21.4 Monitoring Response

Threshold Breach Process: Incident Trigger Criteria: Risk Reassessment Trigger Criteria: Change Trigger Criteria:

22. Assurance Governance

22.1 Assurance Requirements

Assurance Scope: Assurance Frequency: Assurance Owner: Required Independence:

22.2 Assurance Activities

Applicable activities may include:
  • control assessment;
  • compliance assessment;
  • technical assessment;
  • model validation;
  • operational review;
  • independent review;
  • audit;
  • evidence review;
  • management review.
Applicable Assurance Activities:

22.3 Assurance Reporting

Assurance Reporting Authority: Finding Escalation Requirements: Corrective-Action Follow-Up:

23. Incident Governance

23.1 Incident Management

Incident Management Procedure: Incident Repository: Incident Owner:

23.2 Incident Severity

Severity Methodology:

23.3 Incident Escalation

Incidents should be escalated according to:
  • severity;
  • affected stakeholders;
  • potential harm;
  • business impact;
  • regulatory significance;
  • safety impact;
  • security impact;
  • privacy impact;
  • control failure.
Escalation Requirements:

23.4 Incident Closure

Closure Authority: Closure Evidence Requirements: Lessons-Learned Requirements:

24. Change Governance

24.1 Change Requirements

Material AI changes should be assessed before implementation. Change Management Procedure: Change Owner:

24.2 Change Classification

Change Classification Method:

24.3 Change Criteria

Potential material changes include:
  • intended-purpose changes;
  • model changes;
  • model-version changes;
  • training-data changes;
  • inference-data changes;
  • data-pipeline changes;
  • feature changes;
  • integration changes;
  • provider changes;
  • human-oversight changes;
  • control changes;
  • operating-environment changes.
Additional Criteria:

24.4 Change Approval

Change Approval Authority: Required Reviews: Testing Requirements: Rollback Requirements:

24.5 Post-Change Review

Post-Change Monitoring: Post-Change Validation: Closure Requirements:

25. AI Retirement Governance

25.1 Retirement Requirements

AI systems should be formally retired when:
  • the system is no longer required;
  • the system is replaced;
  • risk becomes unacceptable;
  • required controls can no longer be maintained;
  • the system becomes obsolete;
  • the intended purpose ends;
  • another approved retirement trigger occurs.

25.2 Retirement Approval

Retirement Approval Authority:

25.3 Retirement Activities

  • authorization;
  • controlled shutdown;
  • access removal;
  • data disposition;
  • dependency closure;
  • evidence retention;
  • residual-risk closure;
  • monitoring closure;
  • stakeholder communication.

25.4 Retirement Evidence

Required evidence may include:
  • retirement approval;
  • shutdown record;
  • access-removal record;
  • data-disposition record;
  • dependency-closure record;
  • risk-closure record;
  • final monitoring record;
  • final validation;
  • lessons learned.

26. Documentation and Records

26.1 Required Records

Governance records may include:
  • AI inventory;
  • AI system profiles;
  • classification records;
  • risk assessments;
  • risk acceptance records;
  • treatment plans;
  • control assessments;
  • approvals;
  • deployment records;
  • change records;
  • incident records;
  • monitoring records;
  • assurance records;
  • management reviews;
  • improvement records;
  • retirement records.

26.2 Evidence Traceability

Material records should be traceable through:

26.3 Record Retention

Retention Requirements: Records Owner: Records Repository: Access Requirements: Disposition Requirements:

27. Reporting

27.1 Governance Reporting

Reporting Frequency: Recipients: Reporting Owner:

27.2 Reporting Content

Reporting may include:
  • AI inventory status;
  • classification status;
  • AI risk profile;
  • residual risks;
  • control effectiveness;
  • incidents;
  • assurance findings;
  • accepted risks;
  • material changes;
  • monitoring results;
  • exceptions;
  • improvement actions;
  • retirement activity.
Additional Reporting Content:

28. Governance Metrics

Organizations may establish governance metrics. Additional Metrics:

29. Regulatory and External Requirements

29.1 Applicable Requirements

Identify applicable:
  • laws;
  • regulations;
  • standards;
  • contractual requirements;
  • internal policies;
  • external commitments.

29.2 Regulatory Change Monitoring

Monitoring Owner: Review Frequency: Change Escalation Process:

30. Third-Party AI Governance

30.1 Third-Party Services

Third-Party AI Services Within Scope:

30.2 Third-Party Requirements

Requirements may include:
  • due diligence;
  • contractual controls;
  • security requirements;
  • privacy requirements;
  • AI-risk assessment;
  • assurance;
  • monitoring;
  • incident notification;
  • change notification;
  • model-change notification;
  • evidence;
  • termination requirements.
Applicable Requirements:

30.3 Third-Party Monitoring

Monitoring Requirements: Review Frequency:

31. Governance Exceptions

31.1 Exception Request

Exception ID: Requirement: Reason: Risk: Compensating Controls: Requested Duration: Owner: Approval Authority:

31.2 Exception Review

Review Date: Review Outcome: Residual Risk: Additional Actions: Expiry / Renewal Date:

32. Governance Review

The governance arrangement should be reviewed periodically and when material events occur.

32.1 Periodic Review

Review Frequency: Next Review Date: Review Owner:

32.2 Triggered Review

Review triggers may include:
  • significant incidents;
  • material changes;
  • regulatory developments;
  • assurance findings;
  • changes in organizational strategy;
  • changes in AI technology;
  • changes in risk appetite;
  • changes in organizational context;
  • significant stakeholder concerns;
  • continual-improvement findings.
Additional Triggers:

32.3 Review Outputs

Required Outputs:

33. Management Review

33.1 Review Inputs

Management review may consider:
  • AI portfolio;
  • governance performance;
  • risk status;
  • control effectiveness;
  • monitoring;
  • incidents;
  • assurance;
  • exceptions;
  • regulatory developments;
  • changes;
  • stakeholder feedback;
  • improvement actions.
Applicable Inputs:

33.2 Review Decisions

Possible outcomes include:
  • continue;
  • improve;
  • change;
  • restrict;
  • suspend;
  • retire;
  • change governance requirements;
  • change monitoring requirements.
Decision Record:

34. Continuous Improvement

34.1 Improvement Sources

Improvement opportunities may originate from:
  • incidents;
  • monitoring;
  • assurance;
  • audits;
  • risk assessments;
  • control assessments;
  • stakeholder feedback;
  • management reviews;
  • technology developments;
  • regulatory developments;
  • implementation lessons learned.

34.2 Improvement Actions


35. Governance Approval

35.1 Approval Record

Prepared By: Role: Date: Reviewed By: Role: Date: Approved By: Role: Date:

35.2 Approval Decision

Decision:
  • Approved
  • Approved with Conditions
  • Rejected
  • Deferred
  • Suspended
Conditions:

35.3 Approval Evidence

Approval Record ID: Supporting Evidence IDs:

36. Governance Status

Governance Record ID: Current Status: Effective Date: Next Review Date: Governance Owner: Approval Authority: Related AIGO Records:

37. Template Traceability

This template should maintain traceability to relevant AIGO framework components.

38. Template Record Relationships

Where applicable, the completed governance record should link to:

39. Template Completion Review

Before the completed governance record is approved, verify:
  • Scope is defined
  • Context is documented
  • Objectives are defined
  • Governance authority is assigned
  • Roles are assigned
  • Decision rights are defined
  • AI inventory requirements are established
  • Classification requirements are established
  • Risk governance is established
  • Lifecycle governance is established
  • Controls are assigned
  • Human oversight requirements are defined
  • Monitoring requirements are defined
  • Assurance requirements are defined
  • Incident governance is established
  • Change governance is established
  • Retirement governance is established
  • Evidence requirements are established
  • Reporting requirements are established
  • External requirements are identified
  • Third-party governance is addressed
  • Exceptions are governed
  • Review requirements are established
  • Management review is defined
  • Continual improvement is established
  • Approval authority is identified
  • Related records are linked

40. Template Usage Instructions

This template should be completed according to the organization’s approved AIGO governance procedures. The completed governance record should be:
  • controlled;
  • versioned;
  • attributable;
  • approved;
  • periodically reviewed;
  • updated following material changes;
  • supported by appropriate evidence;
  • retained according to applicable requirements.
The template should be adapted to the organization’s:
  • governance structure;
  • organizational context;
  • AI portfolio;
  • risk methodology;
  • classification methodology;
  • control framework;
  • regulatory environment;
  • roles and responsibilities;
  • evidence architecture;
  • records-management requirements.
A completed template should not replace a required policy, procedure, risk assessment, control assessment, approval record, incident record, assurance report, or other controlled artifact. Where another controlled document contains the authoritative requirement, this template should reference that document.

41. Template Governance

41.1 Template Owner

Template Owner:

41.2 Template Review

Review Frequency: Next Review Date:

41.3 Template Change Control

Changes to this template should be managed through the applicable AIGO change-management process. Material changes should consider their effect on:
  • related procedures;
  • templates;
  • schemas;
  • evidence requirements;
  • mappings;
  • tools;
  • completed governance records.

42. Document Control


43. Template Status

Document: AIGO — AI Governance Template Version: 0.2 Status: Draft Working Name: AIGO Full Name: AI Governance Operating Framework Document Identifier: AIGO-TPL-001 Document Type: Governance Template This template provides the baseline structure for establishing, documenting, approving, operating, reviewing, and improving an organization’s AI governance arrangement under the AIGO AI Governance Operating Framework.

44. End of Template

AIGO — AI Governance Template Document ID: AIGO-TPL-001 Version: 0.2 Status: Draft End of Template