Clinical safety
Clinical boundary, escalation, unsafe-output review and human factors.
ASSURANCE MODEL
Governance must scale with the intelligence.
V10 defines the functions an independent or appropriately constituted Disease Intelligence assurance structure should cover. No individuals are presented as appointed members unless formally appointed.
Data governance
Provenance, permissions, minimisation, retention, quality and secondary-use controls.
AI assurance
Model/version control, evaluation, bias, abstention, monitoring and auditability.
Security
Threat modelling, access control, incident response, vulnerability management and supplier risk.
Patient/public interest
Clarity, accessibility, consent, fairness, complaints and meaningful user representation.
Research integrity
Protocol registration, benchmark transparency, conflict-of-interest and publication discipline.