JCAG-A: Extending Judgment-Centered AI Governance to Agentic Systems

This paper introduces JCAG-A, the agentic extension of the Judgment-Centred AI Governance (JCAG) framework, addressing the structural governance gap created when AI systems act autonomously rather than produce recommendations for human review. The paper introduces the Window of Accountability Collapse (WAC) as the failure mode unique to agentic systems, and proposes five governance dimensions.