Skip to content
Compliance & GovernancePublished Nov 13, 2026 · 12 min read

ISO 42001 gap checklist: the 65-point AI management system readiness assessment.

Your customers are starting to ask a new question in security questionnaires and RFPs: how do you govern your AI? ISO/IEC 42001 is the first certifiable international standard that lets you answer with evidence instead of a slide. Most organizations that start an ISO 42001 project discover the same thing within weeks: they already use more AI than anyone had inventoried, and little of it has a documented owner, risk assessment or impact assessment. This checklist turns the standard into 65 practical checks: 27 for the mandatory clauses (4–10) and 38 for the Annex A controls.

Share
Executive summary

An ISO 42001 gap analysis compares your current AI governance against the mandatory requirements in Clauses 4–10 and the 38 Annex A controls you select through your Statement of Applicability. You are audit-ready when your AI inventory, AI policy, AI risk assessment, AI system impact assessments, Statement of Applicability, internal audit and management review are all in place and producing records.

Key takeaways
  • 01ISO/IEC 42001 has two auditable parts: mandatory Clauses 4–10 and 38 Annex A reference controls across nine objectives (A.2–A.10).
  • 02Annex A controls are selected through your AI risk assessment and justified in a Statement of Applicability — not all are mandatory.
  • 03ISO 42001 requires both AI risk assessment and AI system impact assessment on individuals, groups and society.
  • 04ISO 42001 certification does not give a presumption of conformity with the EU AI Act, but it builds the governance those obligations need.
  • 05Organizations with ISO/IEC 27001 can integrate risk, internal audit and management review and move faster.
Definitions

What ISO/IEC 42001 requires

Governance team reviewing AI system monitoring and an ISO 42001 gap checklist

ISO/IEC 42001:2023 specifies the requirements for an Artificial Intelligence Management System (AIMS): the policies, roles, processes and controls an organization uses to develop, provide or use AI responsibly. It follows the same harmonized structure as ISO 9001 and ISO/IEC 27001, so organizations with an existing management system can extend it rather than start over.

  • Clauses 4 to 10: the mandatory requirements that apply to every certified organization.
  • Annex A: 38 reference controls grouped under nine control objectives (A.2 to A.10). You select the controls that treat your AI risks and justify exclusions in a Statement of Applicability (SoA). Annex B gives implementation guidance.

Why a gap checklist matters now

  • Customers and procurement teams increasingly ask vendors that build or embed AI for evidence of AI governance.
  • The EU AI Act timeline is set. Under the Digital Omnibus on AI, obligations for stand-alone (Annex III) high-risk AI systems apply from 2 December 2027, and for AI embedded in regulated products (Annex I) from 2 August 2028. Most transparency obligations still apply from 2 August 2026.
  • Certification is mature. Accredited certification bodies audit against ISO 42001 under ISO/IEC 17021-1 and ISO/IEC 42006:2025, with a three-year certificate cycle and annual surveillance audits.
Checklist · Part A

Management system clauses (27 checks)

Tick only what you can prove with a record or document today.

Clause 4: Context of the organization

✓Internal and external issues relevant to AI are documented, including your role(s) in the AI value chain.
✓Interested parties and their requirements are identified, including customers, regulators and applicable laws such as the EU AI Act.
✓An AI system inventory exists: every AI system and AI-enabled tool in scope, with owner, purpose and role.
✓The AIMS scope is documented, including boundaries and exclusions.
✓The AIMS processes and their interactions are defined.

Clause 5: Leadership

✓Top management demonstrates commitment: resources approved, AIMS objectives aligned with strategy, involvement recorded.
✓An AI policy is approved, communicated and available to interested parties as appropriate.
✓Roles, responsibilities and authorities for the AIMS are assigned and communicated.

Clause 6: Planning

✓An AI risk assessment process is defined, with criteria for acceptable risk.
✓AI risks are assessed and a risk treatment plan is approved by risk owners.
✓A Statement of Applicability lists the selected Annex A controls and justifies every inclusion and exclusion.
✓An AI system impact assessment process exists for consequences to individuals, groups and society.
✓Measurable AI objectives are set, with owners, resources and deadlines.

Clause 7: Support

✓Resources for the AIMS are determined and provided.
✓Competence requirements for AI roles are defined, and evidence of competence is kept.
✓Staff are aware of the AI policy, their contribution and the consequences of not conforming.
✓Internal and external communication on AI is planned.
✓Documented information is controlled: versioning, approval, access and retention.

Clause 8: Operation

✓Operational processes for AI are planned, controlled and documented, including outsourced processes.
✓AI risk assessments are repeated at planned intervals and when significant changes occur.
✓The AI risk treatment plan is implemented and its effectiveness is recorded.
✓AI system impact assessments are performed and retained.

Clause 9: Performance evaluation

✓Monitoring and measurement of the AIMS and AI systems is defined and results are analyzed.
✓An internal audit programme covers the full AIMS, and at least one full audit cycle is complete.
✓Management review has been held, with required inputs and recorded decisions.

Clause 10: Improvement

✓Continual improvement of the AIMS is planned and evidenced.
✓Nonconformities, including AI incidents such as model drift, harmful outputs or misuse, are recorded, corrected and analyzed for root cause.
Checklist · Part B

Annex A controls (38 checks)

Mark each control as in place, partial, missing or not applicable (justified in your SoA).

A.2 Policies related to AI (3)

✓A.2.2 AI policy — documented and approved.
✓A.2.3 Alignment with other organizational policies — linked to security, privacy, quality and ethics policies.
✓A.2.4 Review of the AI policy — reviewed at planned intervals and after significant change.

A.3 Internal organization (2)

✓A.3.2 AI roles and responsibilities — defined across the AI lifecycle.
✓A.3.3 Reporting of concerns — a channel exists for staff to raise AI concerns without reprisal.

A.4 Resources for AI systems (5)

✓A.4.2 Resource documentation — resources for each AI system identified and documented.
✓A.4.3 Data resources — datasets used are documented.
✓A.4.4 Tooling resources — tools, frameworks and platforms are documented.
✓A.4.5 System and computing resources — infrastructure and compute are documented.
✓A.4.6 Human resources — people and competencies needed are documented.

A.5 Assessing impacts of AI systems (4)

✓A.5.2 AI system impact assessment process — defined and triggered at the right points.
✓A.5.3 Documentation of AI system impact assessments — results retained.
✓A.5.4 Assessing impact on individuals or groups — fairness, rights and safety considered.
✓A.5.5 Assessing societal impacts — wider societal effects considered.
Checklist · Part B continued

Lifecycle, data, users and suppliers

A.6 AI system life cycle (9)

✓A.6.1.2 Objectives for responsible development — documented and applied.
✓A.6.1.3 Processes for responsible design and development — defined lifecycle process.
✓A.6.2.2 Requirements and specification — documented for new systems and major changes.
✓A.6.2.3 Documentation of design and development — design decisions recorded.
✓A.6.2.4 Verification and validation — test plans, acceptance criteria and results.
✓A.6.2.5 Deployment — release plan and deployment approval.
✓A.6.2.6 Operation and monitoring — performance, drift and error monitoring in production.
✓A.6.2.7 Technical documentation — available to the right users and parties.
✓A.6.2.8 Recording of event logs — logging enabled and retained.

A.7 Data for AI systems (5)

✓A.7.2 Data for development and enhancement — data management processes defined.
✓A.7.3 Acquisition of data — sources, rights and selection criteria documented.
✓A.7.4 Quality of data — quality requirements defined and checked.
✓A.7.5 Data provenance — origin and lineage recorded.
✓A.7.6 Data preparation — preparation and labelling methods documented.

A.8 Information for interested parties (4)

✓A.8.2 System documentation and information for users — users get what they need to use the system properly.
✓A.8.3 External reporting — interested parties can report adverse impacts.
✓A.8.4 Communication of incidents — an incident communication plan exists.
✓A.8.5 Information for interested parties — obligations to share information are identified and met.

A.9 Use of AI systems (3)

✓A.9.2 Processes for responsible use — defined for staff using AI tools.
✓A.9.3 Objectives for responsible use — documented.
✓A.9.4 Intended use — systems used only as intended, with monitoring.

A.10 Third-party and customer relationships (3)

✓A.10.2 Allocating responsibilities — split between you, partners, suppliers and customers is clear.
✓A.10.3 Suppliers — AI suppliers and third-party models are assessed and monitored.
✓A.10.4 Customers — customer expectations and needs are considered.

Want us to score this with you?

Book a free ISO 42001 gap review call. We will review your AI use, scope and target date and show you where your biggest gaps are.

Book a free gap review →
Scoring

How to score your ISO 42001 readiness

Count the checks that are fully in place, then divide by the number of applicable checks (all 27 clause checks plus the Annex A controls in your SoA).

ISO 42001 readiness score bands
Readiness scoreStatusRecommended next step
85–100%Audit-readyClose remaining minor gaps, complete an internal audit and management review, then book Stage 1
60–84%Close, with gapsRun a formal gap assessment and close actions over 2–4 months
30–59%Foundation in placeFocus on the AI inventory, risk and impact assessments, and SoA over 3–6 months
Under 30%Early stageStart with scope, AI policy, roles and training
Timeline

Typical ISO 42001 implementation timeline

Timelines depend on scope, the number of AI systems and your starting point. These ranges are typical for a mid-sized organization:

01Gap assessment and AI inventory: 2–4 weeks.
02AI policy, roles, risk and impact assessment methods: 1–2 months.
03Control implementation and Statement of Applicability: 2–4 months.
04Internal audit and management review: 1 month.
05Certification audit (Stage 1 and Stage 2) by an accredited certification body: 1–2 months.

Organizations that already hold ISO/IEC 27001 can often move faster, because the risk process, internal audit and management review structure carry over.

Already certified to ISO/IEC 27001? See our ISO 27001 consulting for an integrated approach.

Pitfalls

7 common ISO 42001 gaps we see

  • No complete AI inventory. Shadow AI — staff using public AI tools — is almost always missing.
  • Confusing risk assessment with impact assessment. ISO 42001 requires both: risk to the organization and impact on people and society.
  • A generic AI policy copied from a template, with no link to actual systems or objectives.
  • A weak Statement of Applicability that excludes controls without justification.
  • No supplier oversight for third-party models, APIs and AI-enabled SaaS.
  • No production monitoring for drift, performance degradation or harmful outputs.
  • No completed internal audit before booking the certification audit.
How we help

Getting ISO 42001-ready with Jags Assurance

Jags Assurance brings more than 20 years of quality and management system auditing experience to AI governance. We help organizations across the United States, Canada and the Middle East build an AI management system that passes audit and works in practice.

  • ISO 42001 gap assessment against Clauses 4–10 and all 38 Annex A controls, with a scored gap register and roadmap.
  • AI inventory, risk assessment and impact assessment facilitation.
  • AIMS documentation: AI policy, procedures and Statement of Applicability.
  • Integration with ISO/IEC 27001 and ISO 9001 to avoid duplicated effort.
  • Internal audit and certification readiness before your certification body arrives.
  • PECB ISO/IEC 42001 training — Foundation, Lead Implementer and Lead Auditor — through Jags Assurance Academy.

Build in-house capability with PECB ISO/IEC 42001 training, or explore our ISO 42001 consulting services.

Book a free ISO 42001 gap review call

We will review your AI use, scope and target date and tell you where your biggest gaps are. Consultants prepare you; an accredited certification body issues the certificate, so your audit stays independent.

Book your gap review →
Frequently asked

Questions we get on this topic

What is an ISO 42001 gap analysis?

A structured comparison of your current AI governance against ISO/IEC 42001, covering the mandatory Clauses 4–10 and the applicable Annex A controls. It produces a list of gaps, their severity and a roadmap to certification.

How many controls are in ISO 42001 Annex A?

Annex A contains 38 controls grouped under nine control objectives, numbered A.2 to A.10. Not all are mandatory; you select them based on your AI risk assessment and justify exclusions in your Statement of Applicability.

Is ISO 42001 mandatory?

No. ISO 42001 is voluntary. Organizations adopt it to demonstrate responsible AI governance to customers, partners and regulators.

Does ISO 42001 certification mean EU AI Act compliance?

No. ISO 42001 does not give a presumption of conformity with the EU AI Act. It does build the governance, risk management and documentation practices that support EU AI Act obligations.

How long does ISO 42001 certification take?

For a mid-sized organization, moving from gap assessment to certification commonly takes 4 to 9 months, depending on scope and starting maturity.

Can we combine ISO 42001 with ISO 27001?

Yes. Both follow the same management system structure, so policies, risk processes, internal audit and management review can be integrated. The control sets remain different.

Who can certify us to ISO 42001?

Only an accredited certification body. Consultants, including Jags Assurance, help you prepare but do not issue certificates, which keeps the certification audit independent.

Get in touch

Book a free ISO 42001 gap review call

We will review your AI use, scope and target date and tell you where your biggest gaps are. Consultants prepare you; an accredited certification body issues the certificate, so your audit stays independent.