Responsible artificial intelligence and reliable management
AI governance
Organisations adopting AI need clear responsibilities, human oversight and control of risks across data, models and suppliers.
Questions for decision-makers
- Who is accountable for each AI use?Decisions, approvals and escalation must have named owners.
- What is supervised by people?Human oversight should be defined where outcomes affect people or operations.
- Which suppliers provide models or data?Third-party models and datasets carry their own risks and conditions.
- How is a decision traced?Records should allow an outcome to be reconstructed and reviewed.
Limits that matter
Management of AI is not the same as evaluating a specific model. ISO/IEC 42001 addresses an AI management system; it does not, by itself, attest that every AI system or product is safe or accurate.
Related service
In development
AI governance assessment
Assessment of how an organisation governs its use of AI against defined criteria, including ISO/IEC 42001 as a reference.
Reference sources
Standards are cited as references that are explained. Citing a standard does not mean that IAC has evaluated any organisation against it.