What is overfitting in machine learning?

A. When a model performs well on training data but poorly on new data
B. When a model performs well on both training and new data
C. When a model performs poorly on training data but well on new data
D. When a model fails to identify patterns in the data