Monitoring & Drift
Operational controls for monitoring & drift — with maturity levels, evidence requirements, and implementation guidance.
5 controls
AI Performance Baseline
Establish documented, quantified performance baselines for production AI systems against which ongoing performance can be compared.
Model Drift Detection
Monitor production AI systems for data drift, concept drift, and output distribution shifts that indicate degraded or changed model behavior.
AI Bias and Fairness Monitoring
Continuously monitor AI system outputs for discriminatory patterns across protected demographic attributes in production.
AI Output Anomaly Detection
Automatically detect unusual, unexpected, or potentially harmful AI outputs in production for investigation and response.
Continuous Model Evaluation
Run ongoing evaluation pipelines against held-out test sets and curated adversarial examples to continuously measure model performance in production.