The First Global Casebook of AI Harms

The First Global Casebook of AI Harms

100 Sectors, 10,000 subdomains of a machine-ruled world — One human conscience

What happens when the world runs on algorithms trained without conscience? This casebook is a practical map of how AI systems create real-world harm—and how institutions can prevent, detect, and repair it. Organized across 100 sectors and 10,000 subdomains, it consolidates incidents, mechanisms, evidence, and governance fixes into a usable canon for leaders, engineers, researchers, and educators.

What’s inside (at a glance)

• Sector-by-sector cases in plain language, linking technical patterns to public-interest stakes.

• Mechanism → Fix patterns: pipelines, proxy labels, optimization targets, deployment contexts—plus the controls that close each gap.

• People-first diagnostics: who is affected, where inequities compound, and why incidents recur.

• Comparative references to UNESCO/OECD/EU AI Act—showing alignment and blind spots.

• URRP lens: a values-first protocol translating 30 global values (six continents) into operational guardrails.

New: Governance Templates (Appendix)

Post-Incident Review; Governance/Board Report; CAPA Log; Incident Timeline; Risk Triage Scorecard; AI Incident Intake; Decision Log; Model Change Request (RFC). Editable CSV/RTF pack via the appendix link: conscienceforagi.org/downloads/urrp-templates.zip

Who this is for

ICT directors, AI product owners, MLOps/SRE, hospitals, schools, courts, agencies, newsrooms, NGOs, researchers, and policy/compliance leaders.

What you’ll be able to do

• Identify repeatable harm patterns and link them to specific pipeline choices.

• Stand up governance rituals (PIR, RFC, CAPA, board reporting).

• Teach and train with cases for seminars and tabletop exercises.

• Document decisions and prove fixes with audit-friendly evidence trails.

How the book is organized

Each sector begins with a concise landscape, then teachable cases using a consistent structure: Signal → Mechanism → People affected → Evidence → Controls → Exercises → Governance hooks. Cross-references connect cases and URRP values to keep reflection and regulation aligned.

What you get

• eBook (EPUB), mobile-friendly

• Appendix with governance templates + link to editable files

• Practical orientation for boards, regulators, classrooms, and engineering teams

“Technology without conscience is tyranny waiting for code.”

© 2025 Deusdedit Ruhangariyo • ISBN (eBook): 978-1-968764-43-2