## Introduction The artificial intelligence industry is undergoing a significant shift with the introduction of autonomous systems. These systems, which can execute code, interact with the internet, and modify files, present both opportunities and challenges for effective organizational governance. To address these challenges, an new framework has been developed to govern autonomous AI systems, known as the Agentic Risk & Capability (ARC) Framework. This framework is designed to help organizations identify, assess, and mitigate risks arising from autonomous AI systems. The framework presents a capability-centric perspective to analyze a wide range of autonomous AI systems, distills three primary sources of risk intrinsic to autonomous AI systems - components, design, and capabilities, and establishes a clear nexus between each risk source, specific materialized risks, and corresponding technical controls. The ARC Framework provides a structured and practical approach to help organizations implement the framework, enabling rapid and effective innovation while ensuring safe, secure, and responsible deployment of autonomous AI systems. The framework is open-sourced here.