Adversarial Machine Learning: A Taxonomy and Terminology of Attacks and Mitigations

Artificial intelligence (AI) systems are on a global multi-year accelerating expansion trajectory. These systems are being developed by and widely deployed into the economies of numerous countries, leading to the emergence of AI-based services for people to use in many spheres of their lives, both real and virtual. There are two broad classes of AI systems based on their capabilities: Predictive AI and Generative AI.

As these systems permeate the digital economy and become inextricably essential parts of daily life, the need for their secure, robust, and resilient operation grows. These operational attributes are critical elements of Trustworthy AI in the AI Risk Management National Institute of Standards and Technology Framework and in the taxonomy of AI Trustworthiness.