Inputs intentionally designed to cause a machine learning model to misclassify them into a specific, incorrect category.
Generality: 255