ISO/IEC 42001: The AI Management System Standard
ISO/IEC 42001:2023 establishes requirements for an artificial intelligence management system (AIMS). It applies to organizations that develop, provide, or use AI, helping them organize the policies, responsibilities, and processes behind responsible AI use.
Its focus extends beyond cybersecurity to the wider risks and opportunities of AI. An information security program remains important, while AI governance adds questions about intended use, accountability, impacts, and oversight.
- Identify where AI is used and who is accountable for it.
- Document intended uses, limitations, and affected stakeholders.
- Review risks and impacts when systems or uses change.
Official guidance: ISO/IEC 42001:2023.
Translate AI Risks into Security Work
For an AI-enabled application, document the data it can access, the tools it can call, and the decisions people make using its output. Assess the surrounding application and its permissions as well as the model.
OWASP’s 2025 guidance for LLM applications highlights risks such as prompt injection, disclosure of sensitive information, unsafe handling of generated output, and excessive agency. Use this guidance to shape security reviews and application testing.
An illustrative customer-support assistant should only retrieve information the requester is authorized to see. Tool calls need enforceable permissions, and consequential actions should have appropriate approval checks. Treat generated instructions and retrieved documents as untrusted input.
Technical guidance: OWASP 2025 Top 10 for LLM applications.
Choose the Right Reference for the Work
Management system standards, risk guidance, and technical security references serve different purposes. Use them together according to your scope.
ISO/IEC 42001:2023
A requirements standard for an organizational AI management system. A reference for organizations pursuing AI management system certification.
Explore ISO/IEC 42001ISO/IEC 23894:2023
Guidance for integrating AI-specific risk management into organizational activities. It provides risk guidance rather than the management system requirements of ISO/IEC 42001.
Explore ISO/IEC 23894NIST AI RMF
A voluntary framework for managing risks associated with AI. Use it to structure discussions and practices around trustworthy AI. AI RMF 1.0 is undergoing revision; check NIST for updates.
Explore NIST AI RMFNIST Generative AI Profile
NIST AI 600-1 supplements the AI RMF with risks and suggested actions specific to generative AI. It is guidance, not a certification scheme.
Explore the Generative AI ProfileOWASP LLM Security Guidance
The 2025 Top 10 identifies application security risks relevant to LLM and generative AI systems. Use it as a technical reference alongside your governance program.
Explore OWASP guidanceNIST Cyber AI Profile
NIST IR 8596 is a preliminary draft applying Cybersecurity Framework outcomes to AI-related cybersecurity. Follow the project for updates; the draft is not a finalized standard or certification.
Read the preliminary draftFramework and draft status reviewed 3 November 2026. Refer to the publishers for subsequent revisions and authoritative requirements.
Make AI Accountability Part of Your Risk Program
Start with an inventory of AI uses and a clear owner for each one. Choose references that fit your objectives, then connect risks to actions and evidence.
For Leadership
Set acceptable uses and decision authority. Consider business value, impacts on people, and the oversight needed before adopting a new AI capability.
For Security and Engineering
Review data boundaries, supplier dependencies, and tool permissions. Test realistic failure and misuse scenarios before release and after material changes.
For Risk and Compliance
Keep the rationale behind assessments, approvals, and treatment decisions. Distinguish internal readiness work from independent certification and legal obligations.