Chapter III: High-Risk AI System

Abstract Chapter III of the EU Artificial Intelligence Act establishes the core regulatory framework for high-risk AI systems (HRAIS), introducing extensive obligations for providers, deployers, and other actors across the AI value chain. It defines classification rules, essential requirements, and procedural duties to ensure compliance, transparency, and accountability. The regime adopts a risk-based approach inspired by the New Legislative Framework, combining legislative principles with future harmonized standards. Key provisions address risk management, data governance, technical documentation, human oversight, and cybersecurity, while anticipating normative challenges in translating abstract legal concepts into technical standards. The chapter also outlines conformity assessment procedures, certification, and institutional mechanisms to safeguard health, safety, and fundamental rights within the EU internal market.