Machine learning techniques were originally designed for stationary and benign environments in which the training and test data are assumed to be generated from the same statistical distribution. However, when those models are implemented in the real world, the presence of intelligent and adaptive adversaries may violate that statistical assumption to some degree, depending on the adversary. At the same time, adversarial samples can help identify weaknesses in an ML model, which, in turn, can be used to gain valuable insights on how to enhance the model. Links: Wikipedia, Adversarial Sample Generation, Adversarial Examples Fool both Computer Vision and Time-Limited Humans