ISO/IEC 42001 is the international standard for artificial intelligence management systems. It gives organisations that develop, provide or use AI a structured way to govern AI responsibly: setting policy, assessing risks and impacts, controlling the AI life cycle and improving over time. As AI use grows across UAE government, finance, healthcare, logistics and retail, certification to ISO/IEC 42001 by an accredited body is one way to show customers and regulators that AI is managed with care. For the full certification route, see ISO/IEC 42001 certification in the UAE.
The standard in brief
ISO lists ISO/IEC 42001:2023 as the first edition, published in December 2023 by the joint ISO and IEC committee on artificial intelligence. It is a management system standard, so it follows the same high-level structure as ISO/IEC 27001 and ISO 9001: context, leadership, planning, support, operation, performance evaluation and improvement. It applies to organisations of any size and sector that develop, provide or use AI-based products or services.
Why it matters in the UAE
The UAE Cabinet adopted the National Strategy for Artificial Intelligence 2031 in 2019, with objectives that include improving AI governance and regulation. AI systems also process personal data, so the federal Personal Data Protection Law, and the DIFC and ADGM data protection regimes in those free zones, sit alongside any AI governance work. An ISO/IEC 42001 system helps organisations bring these obligations, their own policies and customer expectations into one managed framework.
Step 1: Define the scope and your AI role
Decide which AI systems, products, services, sites and teams the management system covers. Be clear about your role for each system: are you developing models, providing AI-based services to customers, using AI tools from suppliers, or a mix? The role shapes which risks and controls matter most.
Step 2: Build an AI inventory
List the AI systems in scope, what they do, the data they use, who uses their outputs and which decisions they influence. Many organisations discover AI in places they had not considered, such as recruitment tools, chatbots, fraud detection or analytics features inside existing software.
Step 3: Set policy, objectives and responsibilities
Top management sets an AI policy aligned with the organisation's purpose, assigns roles and responsibilities, and sets measurable objectives for responsible AI. Clear ownership for each AI system prevents gaps between technical, legal and business teams.
Step 4: Assess risks and impacts
ISO/IEC 42001 expects an AI risk assessment and an assessment of the impacts AI systems can have on individuals, groups and society, such as unfair outcomes, lack of transparency, privacy harm, safety concerns or misuse. Controls are then selected and documented, and the reasons for including or excluding controls are recorded.
Step 5: Control the AI life cycle
Put controls in place across design, data acquisition and quality, development, verification and validation, deployment, operation, monitoring and retirement. Human oversight, documentation of intended use, information for users and handling of incidents are typical areas, along with control of third-party AI suppliers.
Step 6: Train people and raise awareness
Developers, product owners, risk and compliance teams and business users each need training suited to their role, from responsible data use to recognising when an AI output needs human review.
Step 7: Monitor, audit and review
Monitor AI system performance and the effectiveness of controls, run internal audits and hold management reviews. Then select an accredited certification body for the stage 1 and stage 2 audits. Certification is followed by surveillance audits and recertification.
Common pitfalls
- Scoping only the newest AI product and missing AI features in existing tools.
- Treating AI governance as a technical project with no business or legal owner.
- Impact assessments written once and never revisited after model changes.
- Relying on AI suppliers without checking their controls or contract terms.
- Policies that staff cannot connect to their daily decisions.
What auditors look for
Certification auditors look for evidence that the system works in practice: an AI inventory that matches reality, risk and impact assessments linked to chosen controls, documented decisions on human oversight, records of monitoring and incidents, competence records, internal audit results and management review outputs. Clear, traceable records make the audit smoother and help your own teams understand how AI is governed.
Integrating with ISO/IEC 27001 and ISO/IEC 27701
Because ISO/IEC 42001 shares the common structure, organisations already certified to ISO/IEC 27001 can extend existing processes for risk, documents, internal audit and management review. Privacy work under ISO/IEC 27701 also links closely to AI data governance. See ISO 27001 certification in the UAE.
Questions to ask before you start
- Which AI systems would customers or regulators most want assurance about?
- Who owns each AI system, and who can approve changes to it?
- Which suppliers provide models, data or AI features we rely on?
- How do we currently detect and respond when an AI output is wrong or harmful?
Honest answers to these questions shape a realistic scope and show where the first controls are needed.
How Sterling helps
Sterling helps UAE organisations scope, design and run AI management systems and prepare for audit by an accredited certification body. Sterling consults and prepares; it does not certify. Every Sterling engagement includes a free Effivity ISO software subscription and free Sterling Academy e-learning for your team, along with complimentary awareness training.
Talk to Sterling
Want help with ISO/IEC 42001 certification? Contact Sterling or call +971 50 584 2597, Monday to Saturday from 8 am to 6 pm. A Sterling consultant replies within one working day.


















