SOURCE-LINKED INTELLIGENCE

AI-GRACE: A Use-Case Operationalization Framework for Agentic AI: From Organizational Objectives and Obligations to Deployment Capabilities and Architecture

arXiv · AI, language, vision and robotics · article · Sep 18, 2026 · UTC

Organizations deploying agentic artificial intelligence must determine more than whether a model is trustworthy; they must establish what to validate, control, and observe for a use case to deliver its intended outcome while meeting applicable obligations. This paper proposes AI-GRACE (Agentic Intelligence-Governance, Risk, Assurance, Controls, and Evidence) as a use-case operationalization framework connecting organizational governance with technical implementation. The proposal draws on professional observations and a purposive synthesis of standards and literature, using design science to f

Read original source ↗ Open in workspace

recordType

paper

region

Global