Model drift

Model drift is a phenomenon where the performance of an AI model degrades over time, requiring adjustments or updating.