Volume 12

Volume 12 - AI Governance and Responsible AI

Earlier volumes explain how to build models, prompts, retrieval systems, agents, protocols, distributed workflows, and enterprise services. Volume 12 asks the harder production question: under what conditions should an organization permit those capabilities to affect people, data, money, rights, or critical operations? You will convert concerns into scenarios, controls, reproducible evidence, and bounded decisions. The chapters expose where assurance breaks: missing populations, weak labels, uncontrolled data copies, model-mediated privilege, stale evidence, paper-only controls, expired exceptions, and usage mistaken for value.