Artificial intelligence — Testing of AI

This document provides technology-agnostic guidance for conducting red teaming assessments on AI systems. The document covers: — Definitions and terminology specific to red teaming in AI contexts — Identification of risks, applicability, objectives and attack vectors — Methodologies for planning and executing red teaming assessments on AI systems, aligned with ISO/IEC 29119 processes as appropriate — Procedures for documenting and reporting red teaming findings — Recommendations for integrating red teaming into the AI system lifecycle This document is applicable to testing all types of AI systems.