Can an organization certify to ISO 42001 without first certifying to ISO 27001?

Yes. The two standards are independent and an organization can pursue ISO 42001 certification without an existing ISO 27001 certification. The Annex SL Harmonized Structure means the AIMS implementation requires the same management-system scaffolding regardless of prior ISO experience. An organization with no prior ISO management system implementation runs a longer implementation because the scaffolding has to be built from scratch. The companion article iso-42001-implementation-guide walks through the implementation timeline for organizations with and without prior ISO experience.

How much of the ISO 27001 Annex A control set transfers to ISO 42001?

The control overlap covers areas where AI systems are an information asset the ISMS already addresses: policies, organization, access control, logging, supplier relationships, system lifecycle. The transfer covers roughly half the ISO 42001 Annex A controls with modifications to extend the existing implementation to the AI-specific risk domain. The remaining ISO 42001 Annex A controls (AI impact assessment, AI-specific data lifecycle, AI system requirements, responsible-use logging extensions, AI-specific disclosure) are additive and require new implementation.

How does the integrated audit save effort compared to separate ISO 27001 and ISO 42001 audits?

The integrated audit covers both standards in a single visit with a single audit team competent in both domains. The shared management-system documentation, the shared internal audit program, and the shared management review cycle reduce the preparation effort and the audit visit time. The cost of the combined audit typically runs 60 to 80 percent of the sum of two separate audits. The savings increase in subsequent surveillance cycles because the certification body has the prior audit's familiarity with the integrated management system.

Does the inspection-layer audit record satisfy both ISO 27001 A.8.15 logging and ISO 42001 A.8.3 responsible-use logging?

The record carries the field set both controls read against. The ISO 27001 control expects user identity, timestamp, action, source, and outcome. The ISO 42001 control expects the AI-specific fields (model and version, policy state, input fingerprint, response classifier outcome) in addition to the information security fields. A single record per AI decision with the inspection-layer schema satisfies both controls. The auditor sampling the records under either standard finds the relevant fields in the same record.

Are there ISO 42001 Annex A controls the inspection layer cannot produce evidence for?

The inspection layer produces operational evidence for the runtime controls (responsible use, data classification, audit records, policy enforcement, information for interested parties for the per-decision context). The management-system controls (AI policy at A.2, internal organization at A.3, the AI impact assessment process at A.5, the AI system requirements process at A.9, the third-party AI provider risk at A.10) require management-system documentation and process records the inspection layer does not produce. The full evidence base combines the inspection-layer records with the management-system documentation.

← Blog

May 26, 2026

ISO 42001 vs ISO 27001: How the AI Management System Layers on Top of Information Security

ISO 42001 and ISO 27001 share the same management-system structure (the Annex SL Harmonized Structure) and a substantial portion of the Annex A control catalog. Organizations with an ISO 27001 certification have a head start on ISO 42001 because the management-system processes transfer with modifications. The two standards address different risk domains: 27001 covers information security risks to confidentiality, integrity, and availability of information assets, while 42001 covers AI-specific risks to fairness, reliability under adversarial pressure, transparency, accountability, and the responsible use of AI systems. This piece walks through the structural overlap, the additive AI-specific controls 42001 introduces, the integration pattern for combined audits, and the inspection-layer architecture that produces evidence under both standards.

ByParminder Singh· Founder & CEO, DeepInspect Inc.

Compliance & Regulationiso-42001iso-27001aimsismsai-compliancecertification

ISO 42001 vs ISO 27001: How the AI Management System Layers on Top of Information Security

ISO 42001 and ISO 27001 share the same management-system structure under the Annex SL Harmonized Structure that ISO uses across its management-system standards. The shared structure means an organization with an ISO 27001 information security management system (ISMS) has the same process scaffolding ISO 42001 requires for the AI management system (AIMS): the scope definition, the leadership commitment, the planning, the support, the operations, the performance evaluation, and the improvement cycle. The two standards differ in the risk domain they address: ISO 27001 covers information security risks to the confidentiality, integrity, and availability of information assets. ISO 42001 covers AI-specific risks to fairness, reliability under adversarial pressure, transparency, accountability, and the responsible use of AI systems.

I want to walk through the structural overlap between the two standards, the additive AI-specific controls ISO 42001 introduces, the integration pattern for combined audits, and the inspection-layer architecture that produces the operational evidence under both standards.

The shared management-system structure

The Annex SL Harmonized Structure runs across the ISO management-system standards. Clauses 1 through 10 have the same titles and the same intent across ISO 9001, ISO 27001, ISO 42001, and the others. The shared clauses are: scope, normative references, terms and definitions, context of the organization, leadership, planning, support, operation, performance evaluation, improvement.

An organization that has implemented Clauses 4 through 10 for ISO 27001 has the same scaffolding for ISO 42001. The scope statement extends to identify the AI systems within the AIMS. The leadership commitment extends to include the AI policy and the AI objectives. The planning process extends to include the AI risk assessment. The support process extends to include the AI-specific competence requirements. The operational planning and control extends to include the AI system lifecycle processes. The performance evaluation includes the AI-specific metrics. The improvement process closes the loop on the AI nonconformities.

The transfer reduces the implementation effort by roughly 30 to 40 percent for an organization that already runs the ISMS. The internal audit team that audits the ISMS audits the AIMS with extensions to the auditor competence. The management review cycle adds the AIMS performance data to the existing review.

The shared Annex A control overlap

Both standards have an Annex A control catalog. ISO 27001:2022 Annex A covers 93 information security controls organized in four themes (organizational, people, physical, technological). ISO 42001 Annex A covers 39 AI-specific controls organized in nine areas.

The control overlap covers areas where AI systems are an information asset the ISMS already addresses. Examples include:

The information security policy under ISO 27001 A.5.1 extends to include the AI policy under ISO 42001 A.2 (AI policies). The organization can run a single policy hierarchy that addresses both the information security and the AI policy requirements.

The roles and responsibilities under ISO 27001 A.5.2 extend to include the AI-specific roles under ISO 42001 A.3 (internal organization). The accountability structure runs as a single matrix.

The access control under ISO 27001 A.5.15 extends to the AI system access under ISO 42001 A.8.2 (responsible use). The same identity-aware access model covers the user access to systems and the user access to AI systems.

The logging and monitoring under ISO 27001 A.8.15 extends to the AI decision logging under ISO 42001 A.8.4 (information for interested parties) and the per-decision audit record requirements. The same logging infrastructure carries the events both standards expect.

The supplier relationships under ISO 27001 A.5.19-A.5.22 extend to the third-party AI provider risk under ISO 42001 A.10 (third-party and customer relationships). The same supplier inventory and risk assessment covers both domains.

The system acquisition and development under ISO 27001 A.8.25-A.8.30 extends to the AI system lifecycle under ISO 42001 A.6 (AI system lifecycle). The same change management and secure development practices cover both domains.

The additive AI-specific controls ISO 42001 introduces

Several ISO 42001 Annex A controls do not have a direct ISO 27001 equivalent because they address AI-specific risk domains.

A.5 (AI impact assessment) requires the organization to assess the potential impacts of AI systems on individuals, groups, and society. The control is AI-specific because the impact analysis covers fairness, reliability under adversarial pressure, transparency, and accountability dimensions the information security analysis does not address.

A.6.2.4 (AI system impact assessment specifically) goes deeper into the impact analysis at the system level. The control requires per-system impact records and feeds the risk treatment under Clause 6.1.2.

A.7 (data for AI systems) addresses the AI-specific data lifecycle: data quality for AI training and operation, data lineage for AI inputs, data provenance for the retrieval corpus, and data minimization at the AI request boundary. The control is AI-specific because the AI system's behavior depends on the data in ways that differ from the information security treatment of data assets.

A.8.3 (responsible use logging) goes beyond the ISO 27001 logging control because the AI decision record carries fields that the information security log does not need: the model and version, the policy state at decision time, the input fingerprint, the response classifier outcome, the human review status if applicable.

A.8.4 (information for interested parties) requires the organization to provide AI-specific information to data subjects, regulators, customers, and the public. The disclosure obligation extends beyond the information security disclosure under ISO 27001 to include the AI-specific transparency obligations under GDPR Article 22 and the EU AI Act.

A.9 (AI system requirements) requires the organization to identify and document the AI system requirements, including the functional, the non-functional, and the responsible-use requirements. The control is AI-specific because the requirements process for AI systems is distinct from the systems-engineering requirements process the ISMS covers.

The integration pattern for combined audits

Organizations that run both ISO 27001 and ISO 42001 typically run a single combined management system with an integrated audit program. The combined approach reduces the audit burden because the certification body audits both standards in a single visit, the internal audit team produces a single audit plan, and the management review covers both AIMS and ISMS performance in a single cycle.

The integration pattern requires alignment in three areas. The scope has to cover both the information assets the ISMS protects and the AI systems the AIMS covers. The risk assessment has to address both information security risks and AI-specific risks, with separate risk treatment for each. The operational controls have to address both domains, with explicit mapping of the controls that satisfy requirements under both standards.

The certification body's audit team has to have competence in both domains. Not every ISO 27001 auditor is qualified to audit ISO 42001, and the certification body has to assign an audit team with the necessary competence. The cost of the combined audit is typically 60 to 80 percent of the sum of two separate audits.

Where the inspection layer produces evidence under both standards

The inspection-layer architecture produces operational evidence the auditor samples for both standards in a single pass.

The per-decision audit record carries the information security fields (user identity, timestamp, action, source, outcome) the ISO 27001 logging control expects. The same record carries the AI-specific fields (model and version, policy state, input fingerprint, response classifier outcome) the ISO 42001 responsible-use logging control expects. A single record series supports both audits.

The policy bundle codifies the access control rules (ISO 27001 A.5.15) and the AI system requirements (ISO 42001 A.9) in the same versioned artifact. The auditor reconstructs the access decision and the AI system requirement that applied to each sampled event from the same record series.

The write-path independence of the inspection layer satisfies the evidence integrity requirements of both standards. The application that calls the AI system cannot modify the records.

The cryptographic integrity signature on each record supports the auditor's evidence integrity verification under both standards. The hash chain pointer catches retroactive modification across the record series.

DeepInspect

This is the gap DeepInspect closes for organizations running combined ISO 27001 and ISO 42001 programs. DeepInspect sits inline between the calling application and any HTTP LLM endpoint. For every request, DeepInspect extracts identity context, applies the policy bundle that codifies both the access control rules and the AI system requirements, classifies the prompt content for data classes the ISMS data classification covers and the AI-specific data classes the AIMS data lifecycle covers, runs the response classifier, and commits the per-decision audit record with the field set both standards' auditors expect.

The architecture produces a single record series that supports both audits. The certification body sampling the records for the ISO 27001 logging control under A.8.15 finds the user identity, timestamp, action, source, and outcome. The same body sampling for the ISO 42001 responsible-use logging under A.8.3 finds the model and version, policy state, input fingerprint, and response classifier outcome. The combined audit runs faster because the evidence is in the same place.

If your organization runs ISO 27001 today and is planning the ISO 42001 implementation, let's talk.

Frequently asked questions

Can an organization certify to ISO 42001 without first certifying to ISO 27001?: Yes. The two standards are independent and an organization can pursue ISO 42001 certification without an existing ISO 27001 certification. The Annex SL Harmonized Structure means the AIMS implementation requires the same management-system scaffolding regardless of prior ISO experience. An organization with no prior ISO management system implementation runs a longer implementation because the scaffolding has to be built from scratch. The companion article iso-42001-implementation-guide walks through the implementation timeline for organizations with and without prior ISO experience.
How much of the ISO 27001 Annex A control set transfers to ISO 42001?: The control overlap covers areas where AI systems are an information asset the ISMS already addresses: policies, organization, access control, logging, supplier relationships, system lifecycle. The transfer covers roughly half the ISO 42001 Annex A controls with modifications to extend the existing implementation to the AI-specific risk domain. The remaining ISO 42001 Annex A controls (AI impact assessment, AI-specific data lifecycle, AI system requirements, responsible-use logging extensions, AI-specific disclosure) are additive and require new implementation.
How does the integrated audit save effort compared to separate ISO 27001 and ISO 42001 audits?: The integrated audit covers both standards in a single visit with a single audit team competent in both domains. The shared management-system documentation, the shared internal audit program, and the shared management review cycle reduce the preparation effort and the audit visit time. The cost of the combined audit typically runs 60 to 80 percent of the sum of two separate audits. The savings increase in subsequent surveillance cycles because the certification body has the prior audit's familiarity with the integrated management system.
Does the inspection-layer audit record satisfy both ISO 27001 A.8.15 logging and ISO 42001 A.8.3 responsible-use logging?: The record carries the field set both controls read against. The ISO 27001 control expects user identity, timestamp, action, source, and outcome. The ISO 42001 control expects the AI-specific fields (model and version, policy state, input fingerprint, response classifier outcome) in addition to the information security fields. A single record per AI decision with the inspection-layer schema satisfies both controls. The auditor sampling the records under either standard finds the relevant fields in the same record.
Are there ISO 42001 Annex A controls the inspection layer cannot produce evidence for?: The inspection layer produces operational evidence for the runtime controls (responsible use, data classification, audit records, policy enforcement, information for interested parties for the per-decision context). The management-system controls (AI policy at A.2, internal organization at A.3, the AI impact assessment process at A.5, the AI system requirements process at A.9, the third-party AI provider risk at A.10) require management-system documentation and process records the inspection layer does not produce. The full evidence base combines the inspection-layer records with the management-system documentation.

← All posts