Presentation synopsis
This completed presentation examined how Nigeria can govern AI systems that influence diagnosis, triage, prognosis, treatment, monitoring or resource allocation without sacrificing patient safety, rights or sustainable health-system development.
It argued that existing legal silos in data protection, medical-device regulation, negligence and procurement are necessary but insufficient when applied separately. The paper proposed a Sustainable Clinical AI Governance Framework built on three rules: lifecycle safety; distributed accountability according to control and capability; and sustainability proportionality.
Operational pillars
Seven pillars translate the framework into practice:
- risk classification;
- data quality and equity;
- safety and human factors;
- explainability and contestability;
- accountability and remedy;
- monitoring and change control; and
- sustainable procurement.
The proposed Nigerian roadmap centres on coordinated regulation, clinical-AI standards, a national register, incident learning, shared technical capacity and staged deployment.