ISO Standards

ISO 42001 vs ISO 27001: What Carries Over

ISO 42001 vs ISO 27001: What Carries Over
In short

ISO 42001 and ISO 27001 share the same management system structure, so leadership, scope, risk method, document control, internal audit, management review and corrective action carry over. What is new in ISO 42001 is AI-specific: AI policy, AI risk and impact assessment, lifecycle controls, data for AI and third-party AI. Build one integrated system where you can.

In this article
  1. ISO 42001 vs ISO 27001: how do they compare?
  2. What carries over from your ISMS?
  3. What is new in ISO 42001?
  4. How do you build both without duplicating work?
  5. Frequently asked questions

ISO 42001 vs ISO 27001 is a question that comes up for every company that already holds an information security certificate and now needs to show AI governance. The good news is that the two standards were written to fit together. If you run a working ISMS, you already have a large share of what an AI management system needs. The rest is about AI itself.

ISO/IEC 27001 is the standard for an information security management system (ISMS), focused on protecting the confidentiality, integrity and availability of information. ISO/IEC 42001:2023 is the standard for an AI management system (AIMS), focused on governing the development, provision and use of AI responsibly. Both are management system standards that can be audited and certified.

ISO 42001 vs ISO 27001: how do they compare?#

PointISO 27001ISO 42001
SubjectInformation securityAI governance and responsible AI
Main risk focusThreats to information assetsRisks from AI systems, including effects on people and society
StructureCommon management system structureThe same common structure
ControlsAnnex A, with a control list for securityIts own annex of AI controls, with implementation guidance
Impact assessmentNot a core requirementAI impact assessment is a central idea
CertificationBy an accredited certification bodyBy an accredited certification body

Think of the relationship in layers. The bottom layer is the management system: policy, planning, support, operation, evaluation and improvement. Above it sit the subject-specific layers: security controls for ISO 27001 and AI controls for ISO 42001. If the bottom layer is shared, adding the second subject is much less work than starting from zero.

What carries over from your ISMS?#

A great deal, because the clauses that run the management system follow the same pattern. If your ISMS works well, you can reuse these parts:

  • Context analysis and the list of interested parties, extended to cover AI-related expectations.
  • Leadership commitment, policy approval and the assignment of roles.
  • The risk assessment method, with AI-specific risks added.
  • Document and record control.
  • Competence, awareness and training processes.
  • Internal audit programme and auditors.
  • Management review, with AI topics added to the agenda.
  • Nonconformity and corrective action handling.
  • Supplier management, with AI suppliers and tools added.
  • Many security controls, such as access control, logging, change management and incident response, which protect AI systems too.

An AI impact assessment deserves special attention. It asks you to think about who is affected by an AI system and how, including people outside your organisation, such as customers, applicants or citizens. This is different from a security risk assessment, which looks at harm to your information. Plan time and the right people for it, including those who understand the business use and the legal setting.

What is new in ISO 42001?#

The new content is about AI. You will need an AI policy, a clear inventory of AI systems and their owners, and a way to assess AI risks and the impact of AI systems on people. You will manage the AI lifecycle, from design and data selection to testing, deployment, monitoring and retirement. You will control the quality and provenance of data used for AI. You will define how humans oversee AI, how transparency is handled and how problems are reported. You will manage relationships with AI suppliers, customers and other parties in the value chain.

These topics sit beyond the usual scope of an ISMS. A strong security programme does not automatically show that a model is fair, explainable or fit for its purpose, so do not assume that certification to one standard covers the other.

Watch for conflicts of ownership. Security teams often lead the ISMS, while AI work sits with data science, product or engineering teams. An integrated system needs a clear owner for each AI system and a forum where security, data, legal and business people meet. A short, regular AI governance meeting often does this job.

How do you build both without duplicating work?#

  1. Treat the AIMS as an extension of your management system, not a separate project.
  2. Combine shared processes: one document control method, one internal audit programme, one management review.
  3. Add AI risks to the existing risk register, and add an AI impact assessment step for relevant systems.
  4. Write the AI policy and procedures, and link them to existing security policies.
  5. Map controls so that one control can serve both standards where it fits.
  6. Train staff on both, so people hear one message.
  7. Ask your certification body whether it can audit both together, which may save audit time.

If you do not hold ISO 27001, you can still implement ISO 42001. You will build the shared management system parts as you go, and many of them will later serve ISO 27001 as well.

Our ISO 42001 toolkit offers editable template documents and a gap assessment checklist, with one year of e-mail support and a free upgrade within 6 months. If you want guidance on integrating the two systems, see our ISO 42001 consulting in India. No one can promise an audit result, because the certification body decides.

Finally, plan the order of work. If you are starting from nothing, you may build the shared management system first, then the security controls, then the AI controls. If you already hold ISO 27001, begin with the gap analysis for the AI-specific clauses and controls, and reuse what you have wherever the standard allows.

Frequently asked questions#

Do we need ISO 27001 before ISO 42001?#

No. ISO 42001 can be implemented on its own. Having an ISMS makes the work easier because of the shared structure.

Does ISO 27001 cover AI risk?#

Only partly. It covers information security risks, including those around AI systems and data. It does not address AI-specific issues such as impact on people, so you need an AI management system for that.

Can one audit cover both?#

Many certification bodies offer combined or coordinated audits. Ask yours about scope, audit time and scheduling.

To learn more read our complete guide to ISO 42001 or contact AdWiser to plan your approach.

Found this useful?

Share it with your network, and follow AdWiser Enterprise Consulting on LinkedIn for more certification guides and compliance tips.

Share:

Leave a Comment

Comments are reviewed before they appear. Please keep them relevant and leave out web links.

Ready to put this into practice?

Start with the ISO 42001:2023 Toolkit and adapt it to your organisation, or talk to us about implementation support.

Chat on WhatsApp