Garbage Out - Garbage In: The Rise of Agentic AI and the Crisis of Information Integrity is a timely warning and practical guide for anyone trying to understand what happens when generative and agentic AI systems begin producing, consuming, and trusting their own outputs. Written from the perspective of a seasoned project management and digital transformation professional, the book reframes the familiar "Garbage In, Garbage Out" principle for the AI age, introducing the more urgent danger of "Garbage Out becomes Garbage In"—a recursive cycle in which hallucinations, synthetic content, and machine-generated assumptions can contaminate the very knowledge systems organizations depend on. Blending plain-language explanation with enterprise-focused insight, this work helps project, program, data, governance, and technology leaders recognize the risks of AI-driven truth decay and begin building the oversight, trust tiers, provenance controls, and governance architectures needed to preserve information integrity before the consequences reach organizational—and potentially civilization-scale—proportions.
Note: This book is meant to be used by project, program and data professionals to develop methodologies to mitigate the contamination of large language models (LLMs) and foundational frameworks. The chapter and section structure are purposely numbered and elaborated to be used as a guide for implementation and continued development of new methods and models. In many cases novel concepts and vocabulary have been created since the current corpus of language around artificial intelligence and agentic systems is still being established.
