What term describes the phenomenon where a model performs well on training data but poorly on unseen test data?

Overfitting

Overfitting occurs when a model learns the noise and details in the training data to the extent that it negatively impacts the performance of the model on new data. It is often characterized by high variance.