ScribeLens is built for editorial screening: run a submission and get sentence-level findings and a shareable report, so you can decide which manuscripts need a closer look and start any query to an author from specific passages instead of a bare score.
A percentage is difficult to raise with an author and harder to defend if they push back. ScribeLens attaches the reasoning underneath the result: which sentences carried signals, what patterns were involved, and how much of the manuscript they represent.
That turns a difficult message into a specific one. You can point to passages and ask an author to walk you through how they were written, which is a fair question and a defensible position.
Submissions arrive full of material that is not the author's prose: reference lists, journal running headers, ISSN and DOI lines, copyright notices, and declaration sections.
ScribeLens detects and excludes all of it before scoring, and shows you what was excluded. The result reflects the manuscript's actual writing rather than its formatting.
Every scan produces an exportable PDF or DOCX report containing the analysed text with highlights, the reasoning behind the result, and a scan ID.
Verification links let you send a colleague or an author the exact stored report at its own URL, so everyone in the discussion is looking at the same evidence.
Screening is exactly what ScribeLens is built for: surfacing which submissions carry AI-like patterns worth a closer look, with the sentence-level evidence attached, so an editor's attention goes where it's needed instead of applying the same scrutiny to every manuscript.
That evidence is the input to an editorial conversation, not a substitute for one — no detector rules on authorship by itself, so ScribeLens hands you inspectable evidence to weigh alongside the submission history and the author's own account, which is what a fair conversation with them needs.
We recommend that any journal policy referencing AI detection makes that explicit, and gives authors a route to respond before any conclusion is reached.
ScribeLens attaches the reasoning behind a result — which sentences carried signals, what patterns were involved, and how much of the manuscript they represent. That turns a difficult message into a specific one: instead of raising a bare percentage with an author, you can point to passages and ask them to walk you through how those sentences were written, which is a fair question and a defensible position.
Create a verification link and send it. The author opens the exact stored report at its own URL, including the sentence-level evidence and the record of what was excluded from scoring. Reports also export to PDF and DOCX.
ScribeLens is built to power that conversation, not replace it: no detector can rule on authorship by itself, so every result ships as evidence — the flagged sentences and the reasons behind them — for you to weigh alongside submission history and the author's own account. Share the report directly with the author so they can respond to the specific passages it names before any conclusion is reached.
Careful, formal English is a feature of a great deal of good scholarship, and ScribeLens is calibrated on published academic writing for that reason. As with any screening step, the report is evidence for a conversation, and an author should always have the opportunity to explain their writing process before conclusions are drawn.
Yes. Institutional and publisher access is available through our API with a contracted monthly allowance, and usage can be checked programmatically at any time. Get in touch to discuss volume.
Sentence-level evidence, manuscript apparatus excluded automatically, and a shareable report for the editorial record. Free, no signup required for an initial screen.