AI ethics is widely discussed but rarely operationalised, with organisations publishing principles that collapse the moment a regulator or auditor asks for evidence.

This guide treats ethics not as a values statement but as an auditable operating discipline, linking each pillar to a control, an owner and an evidence trail.

It identifies three common failure patterns: the poster-on-the-wall charter, principles without owners, and ethics committees detached from operational reality.

Eight ethical pillars, including fairness, transparency, accountability and privacy, are each mapped to ISO 42001 clauses and EU AI Act articles.

Written for AI governance leads, AIMS managers and compliance officers, it positions the MENA region at an inflection point where evidenced ethics is becoming a certification and procurement requirement.