Organizational Debt

Why your AI projects fail
1. Auflage 2027
ISBN: 978-1-041-43233-3
Verlag: Taylor & Francis Ltd

Buch, Englisch, 296 Seiten, Format (B × H): 178 mm x 254 mm

Reihe: Security, Audit and Leadership Series

Organizational Debt and AI Failure: Why Your AI Project Fails challenges the assumption that unsuccessful artificial intelligence initiatives are primarily caused by weak models, insufficient computing power, or poor technology choices. Instead, Scott Andersen argues that AI often fails because it is introduced into organizations carrying years of unresolved structural debt.

The book presents a practical framework for identifying and managing five interconnected forms of organizational debt: technical, data, personnel, security, and embedded debt. Technical debt creates fragile architectures and brittle integrations. Data debt undermines meaning, lineage, context, and trust. Personnel debt appears through skill gaps, unclear responsibilities, burnout, and dependence on a few critical individuals. Security debt weakens identity, governance, assurance, and accountability. Embedded debt connects them through undocumented assumptions, informal processes, workarounds, and hidden dependencies.

AI does not necessarily create these weaknesses—it exposes and accelerates them. Processes that remained stable through human judgment and manual intervention can collapse when automated systems begin operating continuously and at machine speed. What appeared to be resilience may have depended on people quietly compensating for organizational problems.

Drawing on decades of enterprise architecture and technology experience, Andersen explains why promising pilots become trapped in “pilot purgatory,” why local improvements can increase systemwide risk, and why failures frequently emerge at the seams between technologies, teams, data, and decision authorities. He introduces practical concepts and measures—including Total AI Debt Exposure and Total AI Operational Pressure—to help leaders make risk visible and actionable.

This book is written for executives, architects, technology leaders, security professionals, data teams, program managers, and organizations preparing to scale AI. Its central message is direct: successful AI transformation does not begin with adoption. It begins with organizational truth. By identifying, measuring, and governing the debt already present, organizations can move beyond experimentation and build AI systems that achieve delivery, adoption, and sustained operational value.

Zielgruppe

Professional Practice & Development, Professional Reference, and Professional Training

Autoren/Hrsg.

Andersen, Scott O.

Fachgebiete

• Mathematik | Informatik EDV | Informatik Informatik Künstliche Intelligenz
• Wirtschaftswissenschaften Betriebswirtschaft Bereichsspezifisches Management Management: Führung & Motivation
• Wirtschaftswissenschaften Finanzsektor & Finanzdienstleistungen Finanzsektor & Finanzdienstleistungen: Allgemeines
• Interdisziplinäres Wissenschaften Wissenschaften: Forschung und Information Risikobewertung, Risikotheorie
• Wirtschaftswissenschaften Betriebswirtschaft Management Risikomanagement
• Wirtschaftswissenschaften Finanzsektor & Finanzdienstleistungen Versicherungswirtschaft

Inhaltsverzeichnis

Part I: Managing Organizational Debt in AI Systems. Chapter 1: From Observation to Diagnosis. Chapter 2: Debt Is the Right Model. Chapter 3: The Four Debts as a System, not a List. Part II: The Hidden Layer. Chapter 4: The Hidden Layer of Organizational Debt. Chapter 5: Technical Debt in the AI Era: From Hidden Friction to Structural Limit. Chapter 6: Data Debt Is Not a Storage Problem. Chapter 7: The impact on Leadership. Chapter 8: Embedded in Security Debt: Silent Catastrophe. Chapter 9: Cross-Cutting Embedded Debts. Part III: When AI Hits Fragile Systems. Chapter 10: Technical Debt as an AI Multiplier. Chapter 11: The Technical Debt Maturity Curve. Chapter 12: Paying Down Technical Debt Without Stopping the Business. Part IV: When AI Loses Ground Truth. Chapter 13: Obscure Data and the Illusion of Completeness. Chapter 14: Level 1: Untrusted, Inconsistent Truth. Chapter 15: Designing Data for AI, Not Dashboards. Part V: When AI Outpaces Human Capacity. Chapter 16: Personnel Debt Is a Systems Problem. Chapter 17: Level 1: Key Person Risk. Chapter 18: The HDMI Operating Model. Chapter 19: AI Changes the Nature of Risk. Chapter 20: The Security Debt Maturity Curve. Chapter 21: Trust Begins in Silicon. Chapter 22: Debt Compounds at the Seams. Chapter 23: Debt as Stored Energy, Not Moral Failure. Chapter 24: Why Traditional Metrics Fail Under AI Pressure. Chapter 25: The Durability Gap: From Pilot to Permanence. Chapter 26: Governance as the Stewardship of Autonomy. The AI Debt Calculator: A Quantitative Framework for AI Readiness and Debt Remediation. Conclusion: The Reckoning You Choose. Appendix 1.

Über Autor(innen)

Scott O. Andersen is a Distinguished Architect, AI Strategist, and Enterprise Transformation Leader with nearly three decades of experience shaping large-scale technology modernization across Fortune 500 enterprises and U.S. federal agencies.

Over the course of his career, Scott has influenced more than $7 billion in technology revenue and mission outcomes across organizations including Microsoft, Lockheed Martin, Verizon, Accenture, CGI, and CSRA/GDIT. His work spans AI architecture, cloud transformation, and mission-critical network design, with a consistent focus on translating advanced technical capability into operational systems that perform under real-world constraints.

Scott is widely recognized for his ability to connect strategy to execution. His work centers on the intersection of AI, infrastructure, and organizational readiness, helping institutions move beyond isolated innovation toward scalable, governed, and durable transformation. He has advised senior leadership across federal agencies including the FAA, DHS, DoD, DOE, FCC, OPM, and others on the impact of AI, Zero Trust architecture, and modern cloud platforms.

At Verizon, where he serves as a Distinguished Architect and Federal Cloud Advisor, Scott has contributed to more than $2 billion in opportunities, while designing AI-driven network and cloud architectures that support both national missions and enterprise-scale operations. He is also the founder of Verizon’s AI-focused newsletter and was recognized as a 2024 Distinguished Communicator for his ability to translate complex technical concepts into actionable insight.

Scott’s career includes foundational contributions to enterprise computing. During his tenure at Microsoft, he led some of the largest enterprise migrations in history, including landmark Domino-to-Exchange transitions, and helped design platforms that generated over $1 billion in revenue. At Lockheed Martin, he led cloud strategy for billion-dollar portfolios, including the development of the Intelligent Cloud Broker, a platform that created more than $100 million in new market revenue.

Beyond his technical leadership, Scott is an active voice in the evolution of AI and infrastructure. He is a frequent speaker at industry events, government forums, and technology conferences, and has appeared on multiple podcasts and media platforms discussing AI, network transformation, and the future of enterprise systems. His thought leadership focuses on the realities of deploying AI at scale, particularly the organizational and infrastructural constraints that determine success or failure.

Scott is also an accomplished author. His published works include There Are Days You Just Want to Go Home, Transitional Services, Danny and the Corporate Ladder, and Infrastructure Architecture for AI Networks. He is currently completing his forthcoming book, Designing AI Data Centers, which explores the convergence of AI workloads, power systems, cooling architectures, and infrastructure design in the era of machine-scale computation.

A recognized leader in his field, Scott holds the Microsoft Certified Architect credential, one of the most elite technical certifications globally, along with multiple additional certifications across cloud, cybersecurity, and enterprise architecture.

His work continues to focus on a central question that defines the future of AI-enabled organizations:

Where should humans govern, where should machines execute, and where must systems be co-designed to sustain both at scale.