Answer
Model inversion attacks attempt to reconstruct sensitive information about the original training data by carefully analyzing a model's outputs.
Real-world example A model inversion attack might try to reconstruct a recognizable face image from a facial recognition model without ever having direct access to the original training photos.