Adversarial machine learning updated

This document provides an overview of adversarial machine learning, including what it is, why it occurs, examples of its effects, and potential mitigations. It begins with introductions to artificial intelligence and machine learning. Adversarial machine learning refers to intentionally introducing perturbations to inputs to cause machine learning models to make mistakes. This can impact applications involving audio, video, images, smart homes, and code. The document discusses the lifecycle of adversarial machine learning attacks and possible defenses against such attacks.