Active
AI Monitoring
The organization shall monitor AI systems to detect security threats, operational issues, and performance anomalies. Monitoring shall include monitoring of model behavior, detection of security incidents, and identification of operational issues, with results reported to relevant stakeholders.