An AI auditor reviewing a model's post-deployment monitoring program finds that alerts are configured only for API error rates and latency. What significant monitoring gap does this reveal?
Select an answer to reveal the explanation.
Short Explanation
You've got a monitoring gap because alerts only for API errors and latency show infrastructure health, not model quality. You need model performance metrics and data drift detection to catch accuracy degradation and distribution shifts.
Full Explanation
The gap is the absence of model performance and data drift monitoring to detect accuracy degradation and distribution shifts. API error and latency alerts show infrastructure health but miss model quality: whether predictions stay accurate and input distributions shift. Comprehensive AI monitoring needs performance metrics and drift detection. Encryption monitoring is a security gap. Data scientist productivity tracking and competitor benchmarking are not standard model monitoring requirements.