The board is asking. The auditors are coming. Your insurance carrier just added an AI exclusion. You don't have a governance program - and the gap between "we're using AI" and "we can prove we're using it responsibly" is widening fast.
Between 55 and 58 percent of organizations are using AI. Fewer than 25 percent have a governance program. The Air Canada chatbot ruling, the Samsung code leak, the Replit production-database wipe, the Arup $25M deepfake fraud - the case law and incident library are no longer hypothetical. Insurance carriers have begun pulling AI coverage from organizations that cannot document their controls.
AI Governance for Practitioners is the actionable playbook for the people who actually have to write, defend, and operate AI policy in 2026. It is not theory. It is not a survey of every framework ever written. It is a concrete program design that a CISO, compliance officer, risk manager, CPO, or SMB executive can start implementing on Monday morning.
What's inside:
- The governance landscape (Part I). Why AI governance matters now, the Big Three (EU AI Act, NIST AI RMF, ISO/IEC 42001), the international landscape (CoE Framework Convention, Singapore MGF for Agentic AI, Colorado AI Act, OECD), and how to classify and inventory AI risks - including agentic risks and the trustworthiness characteristics that drive them.
- Building your program (Part II). Governance program design from zero, human oversight as a design principle, policy development that people actually follow, vendor and third-party AI risk assessment, audit readiness and ISO 42001 certification, and sector-specific considerations (financial services, healthcare, legal, government).
- Looking forward (Part III). Agentic AI governance, the standards acceleration (NIST AI Agent Standards, OpenID AuthZEN, CSA CSAI Foundation), the insurance reckoning, and a 12-month roadmap.
- Templates that work. A complete AI Acceptable Use Policy, AI Risk Assessment, AI Vendor Assessment Scorecard, Incident Response Plan, framework comparison matrix, decision trees, and a 60+ term glossary - all in the appendices.
11 chapters. 7 appendices. SMB Spotlights and Action Items in every chapter. Designed so an executive can read it and start standing up a program the next week.
Includes coverage of the EU AI Act phased timeline, NIST AI RMF GenAI Profile, ISO 42001 certification pathways, Singapore MGF for agentic AI, and the 2025-2026 standards explosion. Cross-references to Securing Agentic AI (2026 Edition) for the technical depth behind the policies.
