GPT-5.1's Drift Detection Addition Suggests Sophisticated Metrics Monitoring Framework

The addition of "drift" to GPT-5.1's monitoring scope — alongside "risks" — suggests a two-tier metrics integrity framework: risk detection (binary: problem exists or doesn't) and drift detection (continuous: metrics slowly moving away from established baselines). Drift detection is a more sophisticated monitoring approach that catches problems before they become acute, requiring trend analysis rather than threshold-based alerting.