Definition
The practice of systematically evaluating AI systems for performance, robustness, security, and adherence to principles of fairness, transparency, and compliance.
In short: The practice of systematically evaluating AI systems for performance, robustness, security, and adherence to principles of fairness, transparency, and compliance.