A structured governance framework for managing AI risk across your entire organization

Enterprise-Wide AI Risk Management (EW-AiRMTM): A Practical Framework for Managing AI Risks in any Organisation, by Prof. Markus Krebsz, introduces a multi-layered governance framework forged from five years of United Nations participatory AI policy development and over thirty years of risk management experience. EW-AiRMTM augments existing ERM frameworks like ISO 31000, COSO, and the UK Orange Book with AI-specific governance capabilities.

The book presents a risk taxonomy drawn from more than 1,700 categorised AI risks across seven evidence-based domains and 24 sub-domains, alongside over 800 risk mitigation controls organised into four actionable quadrants. Six foundational pillars establish governance prerequisites before AI deployment, while the HAiPECR ethical filter keeps human accountability central to every governance decision.

Readers will also find:

• A resilience framework designed to prepare organisations for AI Black Swan events that cannot be fully anticipated in advance
• Practical guidance on extending existing enterprise risk management structures rather than replacing them with entirely new governance architectures
• A risk taxonomy spanning seven domains and 24 sub-domains built from over 1,700 categorised AI risk scenarios
• The HAiPECR ethical framework, hosted on the OECD.AI tool website, embedding human accountability into every governance decision

Written for Chief Risk Officers, board members, compliance leaders, data scientists, and senior executives making AI deployment decisions, this book delivers a field-tested governance framework that addresses the AI-specific risks existing enterprise risk management structures were never designed to handle.

MARKUS KREBSZ is the creator of the EW-AiRM framework, founding director of the Human-Ai.Institute, and Honorary Professor at the University of Stirling. He led the development of the UNECE Declaration on the technical regulation of products/services with embedded artificial intelligence (ECE/TRADE/486), published by the United Nations and open for signature by government agencies worldwide, and created the HAiPECR™ ethical filter, featured in the OECD.AI Catalogue of Tools & Metrics. He served as a member of the European AI Office’s expert group and of all four working groups that drafted the EU General-Purpose AI Code of Practice, serves on the supervisory council of a European bank since 2017, and is a StandICT.eu Standards Maker and a working group member of IEEE P7007.1, the Ontological Standard for Addressing Risks in AI Systems. With over thirty years of experience across capital markets, banking, structured finance, and FinTech, he has authored or edited more than twenty-five publications in the risk and governance domain. Learn more at www.ewairm.com.