MY ONE BLOGby Manishdai
Data & Analytics

What is Model Drift?

Quick answer

The gradual decline in an AI model's accuracy as real-world data or conditions change from those it was trained on.

Model Drift, explained

A model learns patterns from historical data. When customer behaviour, market conditions, or data sources shift, those patterns may no longer hold. Data drift refers to changes in the inputs a model receives, while concept drift refers to changes in the relationship between inputs and the correct output.

For example, a demand forecasting model trained before a major price change might start predicting sales poorly because customers now behave differently. Fraud models face drift constantly as criminals change tactics. Teams detect drift by monitoring input data distributions and comparing predictions with actual outcomes over time. When performance drops below an agreed threshold, the model is retrained on newer data, adjusted, or replaced. Without monitoring, drift often goes unnoticed until it causes business problems.

Related terms

Machine Learning

A branch of AI in which computers learn patterns from data to make predictions or decisions without being explicitly programmed for each case.

Model Evaluation

The process of testing an AI model against defined tasks and criteria to measure how accurate, reliable, and safe it is.

Predictive Analytics

Using historical data, statistics, and machine learning to forecast future outcomes or behaviors.

Guides that mention Model Drift