Skip to content

ISO/IEC 42001:2023 Crosswalk

This document maps the FSI Agent Governance Framework controls to ISO/IEC 42001:2023 — Artificial Intelligence Management System (AIMS), the international standard for organizational AI governance.

ISO/IEC 42001 is complementary to NIST AI RMF (see NIST AI RMF Crosswalk) — the frameworks address different needs and many FSI organizations implement both. NIST AI RMF supports continuous risk identification and measurement; ISO/IEC 42001 provides a certifiable management-system structure (Plan-Do-Check-Act) that aligns with ISO 27001, ISO 9001, and other ISO management systems FSI firms already operate.

Scope note. This crosswalk maps clauses 4–10 (the certifiable management-system requirements) and Annex A (control objectives) to FSI-AgentGov controls. It is provided as an implementation aid for organizations preparing for ISO/IEC 42001 certification or third-party audit. It is not a conformity assessment and does not substitute for a formal gap analysis by an accredited certification body. Organizations should verify scope statements with their internal audit / compliance teams before relying on this mapping for certification evidence.

How to Use This Crosswalk

Audience Use
Compliance Officer preparing for ISO 42001 certification Identify which FSI controls already provide implementation evidence; flag clauses with Partial or N/A coverage for gap remediation.
Internal Auditor Validate that FSI-AgentGov controls produce the artifacts ISO 42001 auditors expect (policies, risk register, monitoring records, management review minutes).
External Auditor (ISO 42001 certification body) Use the mapping as a roadmap for evidence collection — most clauses are addressed by existing Pillar 2 (Management) and Pillar 3 (Reporting) controls.
AI Governance Lead Pair this crosswalk with the NIST AI RMF Crosswalk for end-to-end coverage of both frameworks.

Coverage Status Legend

Status Meaning
Full One or more FSI-AgentGov controls directly address the clause requirements with documented procedures and evidence.
Partial FSI-AgentGov controls provide partial coverage; additional organization-specific procedures or policies are required.
Framework Addressed at the framework layer (governance principles, operating model) rather than via a specific control.
Org Organization-specific responsibility; FSI-AgentGov provides supporting controls but the AIMS scope, policy, and management review must be authored by the organization.
N/A Clause is administrative or scoping in nature and does not map to a technical control.

Microsoft AI Service Certification Context

Microsoft has achieved ISO/IEC 42001 certification covering the following AI services in scope (per Microsoft Learn — ISO/IEC 42001 offering page, retrieved May 2026):

  • GitHub Copilot
  • Microsoft 365 Copilot
  • Microsoft Copilot Health
  • Microsoft Copilot Studio
  • Microsoft Dragon Copilot
  • Microsoft Foundry
  • Microsoft Security Copilot

Audit reports and scope statements are available through the Microsoft Service Trust Portal.

Caveat for FSI maintainers: Microsoft's certification covers Microsoft's own service operations as the AI provider. It does not transfer to customer governance, customer use cases, or the customer's own AIMS. FSI organizations pursuing their own ISO 42001 certification must build and operate their own management system on top of these certified services. Use vendor certification as input to Control 2.7 (Vendor Risk Management) — not as a substitute for the customer's own AIMS evidence.

Clause 4: Context of the Organization

Clause Requirement FSI Controls / Artifacts Coverage
4.1 Understanding the organization and its context Determine external/internal issues relevant to AIMS purpose Executive Summary, Regulatory Framework, Adoption Roadmap Framework
4.2 Understanding needs and expectations of interested parties Identify stakeholders and their AIMS-relevant requirements Operating Model, RACI Matrix, Role Catalog Framework
4.3 Determining the scope of the AIMS Document AIMS scope boundaries and exclusions Framework Layer (Zone definitions in Zones and Tiers) Org
4.4 AI Management System Establish, implement, maintain, continually improve the AIMS Governance Fundamentals, Governance Cadence Framework

Clause 5: Leadership

Clause Requirement FSI Controls / Artifacts Coverage
5.1 Leadership and commitment Top management demonstrates leadership for the AIMS Executive Summary, Operating Model Framework
5.2 AI Policy Establish and document AI policy Framework Layer + 2.13 — Documentation and Record Keeping Org
5.3 Roles, responsibilities, authorities Assign and communicate AIMS roles Operating Model, RACI Matrix, Role Catalog, 1.2 — Agent Registry Full

Clause 6: Planning

Clause Requirement FSI Controls / Artifacts Coverage
6.1.1 Actions to address risks and opportunities Plan AIMS actions for risks/opportunities 2.6 — Model Risk Management, Adoption Roadmap Full
6.1.2 AI risk assessment Establish, implement, maintain AI risk assessment process 2.6, 2.11 — Bias Testing, 2.20 — Adversarial Testing / Red Team Full
6.1.3 AI risk treatment Risk treatment plan with documented controls 2.6, Pillar 1 (Security) controls catalog Full
6.1.4 AI system impact assessment Assess potential impacts on individuals, groups, society 2.11, 2.19 — Customer AI Disclosure Partial
6.2 AI objectives and planning to achieve them Establish measurable AIMS objectives Adoption Roadmap, Governance Cadence Org
6.3 Planning of changes Plan changes to the AIMS in a controlled manner 2.3 — Change Management Full

Clause 7: Support

Clause Requirement FSI Controls / Artifacts Coverage
7.1 Resources Determine and provide AIMS resources Org responsibility; supported by Adoption Roadmap Org
7.2 Competence Ensure personnel competence; retain evidence 2.14 — Training and Awareness Program Full
7.3 Awareness Awareness of AI policy, contributions, implications 2.14, 2.19 — Customer AI Disclosure Full
7.4 Communication Internal/external communications relevant to AIMS 2.13 — Documentation and Record Keeping, 3.3 — Compliance and Regulatory Reporting Full
7.5 Documented information Create, control, retain documented information 1.7 — Comprehensive Audit Logging, 2.13, 1.19 — eDiscovery for Agent Interactions Full

Clause 8: Operation

Clause Requirement FSI Controls / Artifacts Coverage
8.1 Operational planning and control Plan, implement, control AIMS processes 2.1 — Managed Environments, 2.15 — Environment Routing Full
8.2 AI risk assessment Operate the risk assessment process at planned intervals 2.6, Governance Cadence Full
8.3 AI risk treatment Implement the risk treatment plan Pillar 1 (Security) catalog, 2.6 Full
8.4 AI system impact assessment Perform impact assessment for AI systems in scope 2.11, 2.19, 2.21 — AI Marketing Claims and Substantiation Partial

Clause 9: Performance Evaluation

Clause Requirement FSI Controls / Artifacts Coverage
9.1 Monitoring, measurement, analysis, evaluation Determine what to monitor, methods, frequency 2.9 — Agent Performance Monitoring, 3.2 — Usage Analytics, 3.7 — PPAC Security Posture Assessment Full
9.2 Internal audit Conduct planned internal audits of the AIMS 3.3 — Compliance and Regulatory Reporting, Governance Cadence Full
9.3 Management review Top management reviews the AIMS at planned intervals Governance Cadence, Operating Model Framework

Clause 10: Improvement

Clause Requirement FSI Controls / Artifacts Coverage
10.1 Continual improvement Continually improve AIMS suitability and effectiveness Governance Cadence, Adoption Roadmap Framework
10.2 Nonconformity and corrective action React to nonconformities; document corrective action 3.4 — Incident Reporting and Root Cause Analysis, 2.3 — Change Management Full

Annex A — AI Management System Controls

ISO/IEC 42001 Annex A defines control objectives organized into nine categories (A.2 through A.10). The mapping below covers the categories most directly addressed by the FSI-AgentGov catalog. Annex A is informative in the standard — organizations document which Annex A controls they apply via a Statement of Applicability (SoA).

Control Objective FSI Controls / Artifacts Coverage
A.2.2 AI policy documented and approved Org responsibility; supported by Operating Model Org
A.2.3 Alignment with other organizational policies Regulatory Framework, Regulatory Mappings Framework
A.2.4 Review of AI policy Governance Cadence Framework

A.3 — Internal Organization

Control Objective FSI Controls / Artifacts Coverage
A.3.2 AI roles and responsibilities Operating Model, RACI Matrix, Role Catalog Full
A.3.3 Reporting of concerns 3.4 — Incident Reporting Full

A.4 — Resources for AI Systems

Control Objective FSI Controls / Artifacts Coverage
A.4.2 Resource documentation 3.1 — Agent Inventory and Metadata Management, 1.2 — Agent Registry Full
A.4.3 Data resources 1.6 — DSPM for AI, 4.6 — Grounding Scope Governance, 4.7 — M365 Copilot Data Governance Full
A.4.4 Tooling resources 2.5 — Testing, Validation, and Quality Assurance Full
A.4.5 System and computing resources 1.20 — Network Isolation, 2.4 — Business Continuity and Disaster Recovery Full
A.4.6 Human resources 2.14 — Training and Awareness Program Full

A.5 — Assessing Impacts of AI Systems

Control Objective FSI Controls / Artifacts Coverage
A.5.2 AI system impact assessment process 2.11 — Bias Testing and Fairness Assessment, 2.6 Partial
A.5.3 Documentation of AI system impact assessment 2.13 — Documentation and Record Keeping Full
A.5.4 Assessing AI system impact on individuals or groups 2.11, 2.19 — Customer AI Disclosure Partial
A.5.5 Assessing societal impacts Org responsibility; informed by Regulatory Framework Org

A.6 — AI System Lifecycle

Control Objective FSI Controls / Artifacts Coverage
A.6.1.2 Objectives for responsible development Operating Model, Adoption Roadmap Framework
A.6.1.3 Processes for responsible AI design and development 2.5 — Testing, Validation, and Quality Assurance, 2.11 Full
A.6.2.2 AI system requirements and specification 1.2 — Agent Registry, 3.1 — Agent Inventory Full
A.6.2.3 Documentation of AI system design and development 2.13 Full
A.6.2.4 AI system verification and validation 2.5, 2.20 — Adversarial Testing / Red Team Full
A.6.2.5 AI system deployment 2.1 — Managed Environments, 2.3 — Change Management, 2.15 — Environment Routing Full
A.6.2.6 AI system operation and monitoring 2.9 — Agent Performance Monitoring, 3.2 — Usage Analytics Full
A.6.2.7 AI system technical documentation 2.13, 2.19 Full
A.6.2.8 AI system event logs 1.7 — Comprehensive Audit Logging, 3.9 — Microsoft Sentinel Integration Full

A.7 — Data for AI Systems

Control Objective FSI Controls / Artifacts Coverage
A.7.2 Data for development and enhancement of AI systems 1.6 — DSPM for AI, 2.16 — RAG Source Integrity Validation Full
A.7.3 Acquisition of data 4.6 — Grounding Scope Governance, 2.7 — Vendor Risk Management Full
A.7.4 Quality of data for AI systems 2.16, 4.6 Full
A.7.5 Data provenance 2.16, 1.7 — Comprehensive Audit Logging Full
A.7.6 Data preparation 4.7 — M365 Copilot Data Governance, 1.13 — Sensitive Information Types Full

A.8 — Information for Interested Parties

Control Objective FSI Controls / Artifacts Coverage
A.8.2 System documentation and information for users 2.19 — Customer AI Disclosure and Transparency, 2.23 — User Consent and AI Disclosure Enforcement Full
A.8.3 External reporting 3.3 — Compliance and Regulatory Reporting, 3.4 — Incident Reporting Full
A.8.4 Communication of incidents 3.4, 1.21 — Adversarial Input Logging Full
A.8.5 Information for interested parties 2.19, 2.21 — AI Marketing Claims and Substantiation Full

A.9 — Use of AI Systems

Control Objective FSI Controls / Artifacts Coverage
A.9.2 Processes for responsible use of AI systems 2.12 — Supervision and Oversight (FINRA Rule 3110), 2.14 — Training and Awareness Full
A.9.3 Objectives for responsible use of AI systems Operating Model, Governance Fundamentals Framework
A.9.4 Intended use of the AI system 1.14 — Data Minimization and Agent Scope Control, 2.17 — Multi-Agent Orchestration Limits Full

A.10 — Third-Party and Customer Relationships

Control Objective FSI Controls / Artifacts Coverage
A.10.2 Allocating responsibilities 2.7 — Vendor and Third-Party Risk Management Full
A.10.3 Suppliers 2.7 Full
A.10.4 Customers 2.19 — Customer AI Disclosure, 2.23 — User Consent and AI Disclosure Enforcement Full

Coverage Summary

Coverage Count (Approx.) Notes
Full 22 Direct mapping to documented controls
Partial 4 Supported but require organization-specific augmentation
Framework 9 Addressed at framework / governance-principles layer
Org 6 Organization-specific responsibility (AIMS scope, policy, management review)

Interpretation. A high Full count indicates that a customer pursuing ISO/IEC 42001 certification can use FSI-AgentGov as a substantial implementation backbone for an M365 / Power Platform AI estate. The Org-coverage clauses (4.3 scope, 5.2 AI policy, 6.2 objectives, 7.1 resources, 9.3 management review, A.2.2 AI policy approval) are deliberately customer-owned — no external framework can author them on the organization's behalf. The Partial-coverage rows generally require domain-specific impact-assessment language that the organization writes against its own products and customer base.

Relationship to NIST AI RMF

NIST AI RMF and ISO/IEC 42001 are complementary, not competitive:

Aspect NIST AI RMF ISO/IEC 42001
Nature Voluntary risk-management framework Certifiable management-system standard
Structure 4 functions (Govern, Map, Measure, Manage), 72 subcategories PDCA management system (Plan-Do-Check-Act), Annex A control objectives
Certification Not certifiable Third-party certifiable by accredited bodies
Geography U.S.-focused International
Primary value Continuous risk identification and measurement Formal governance structure and certifiable evidence

Recommended sequencing for FSI organizations:

  1. Begin with NIST AI RMF risk assessments (lower implementation barrier) — see NIST AI RMF Crosswalk
  2. Formalize findings into ISO/IEC 42001 AIMS policies and Annex A control mapping
  3. Pursue ISO/IEC 42001 certification (typically 12–18 month engagement with an accredited certification body)
  4. Use NIST AI RMF for continuous risk monitoring between annual ISO 42001 surveillance audits

References

Updated: May 2026 | Version: v1.6.2 | ISO/IEC 42001 Mapping Last Verified: May 2026