Autonomous Audit of AI Governance Architecture

Autonomous Audit of AI Governance Architecture proposes Examination Validity, a governance property that enables an independent examiner to derive sound, verifiable conclusions from any conforming fragment of a distributed AI governance record, even when complete evidence is unavailable. The paper extends runtime governance beyond operational control to autonomous, cross-organizational auditability, demonstrating how cryptographically verifiable evidence, provenance, and governance attestations can support trustworthy investigations across organizational boundaries without introducing new governance layers.