A financial institution deploys a credit-scoring AI model. Six months after deployment, the model's Gini coefficient drops from 0.62 to 0.44. The risk team suspects the model has degraded but is unsure of the cause. Which action should be taken FIRST?

Diagnosing the root cause of performance degradation—data drift, concept drift, or sample composition shift (population shift)—is necessary before selecting the appropriate remediation. Retraining without diagnosis may reproduce the same problem. The immediate retraining option with the most recent 12 months of data is premature and could perpetuate bias if the data has shifted. The revert-to-champion-and-notify-regulator option may be warranted eventually but not before diagnosis. The threshold-increase option addresses symptoms rather than the underlying cause.