62. AI governance and compliance
Set ownership, approval, documentation, audit, risk classification, human oversight, and change-control rules for ML systems. Use model cards, data documentation, impact assessments, evaluation records, and applicable legal or sector requirements.