Is the report admissible if I have to defend a kill decision?

It’s a contemporaneous record, the score, the confidence band, the flagged passages, and your own note, timestamped to your account. It documents the basis for your call. Treat it as decision-support, not legal proof: detection is probabilistic, so the report backs a human judgment rather than replacing it.

What happens to the contributor’s draft after I scan?

It’s saved privately to your account so you can reopen the report, never public unless you create a share link. Run the scan in ephemeral mode to store nothing at all. We don’t use submissions to train detection models.

How do I handle the contributor relationship after a high score?

As a conversation, not a verdict. Share the report, show the flagged passages, and ask. A high score is a reason to talk, not grounds to accuse, false positives happen, and the evidence is there to ground the discussion, not end it.

For editors

For editors reviewing freelance submissions

Score, signals, and a notes pane you can attach to a decision.

You decide what runs under the masthead. Detecto reads a freelance draft and returns an AI-likelihood with the signals behind it, so the desk can weigh a submission on evidence, not a hunch, before it’s scheduled. The report saves privately to your account; share it only when a conversation with the contributor calls for it.

See sample report How calibration works

Built for the people who review other people’s content

Editors at independent magazines
University writing centers
Newsroom verification desks
PR review teams
Content QA leads at agencies

What you review

What editors typically scan with Detecto.

Pitch decks and final drafts from freelancers and contributors
Long-form essays and analysis pieces
Op-eds and reader submissions
Translated copy from non-staff translators

What’s at risk

The failure modes Detecto is built to surface.

A contributor passes off AI-drafted copy as original reporting
A lightly edited AI draft slips through on deadline
Machine-generated text hides inside translated copy
A trusted byline quietly starts leaning on a model

Where Detecto fits

The workflow, end to end.

Five steps. None of them is automated.

01
Paste the draft, or upload the PDF / DOCX export.
02
Read the score and confidence band before you read for line edits.
03
Open the highlighted view to see which passages moved the score.
04
Attach a note recording what you decided and why.
05
Share the report with the contributor only if the conversation needs it.

What the report gives you

Evidence that goes in the file.

A 0–100 AI-likelihood with a High / Medium / Low confidence band.

Sentence-level highlighting that points to the passages in question.

A timestamped, account-scoped record you can attach to a run-or-kill decision.

What not to do with AI detection

The same posture every Detecto reviewer reads. The score is evidence, not a verdict.

Don’t treat a high score as proof, it’s a likelihood, not a confession.
Don’t spike a piece on the number alone; pair it with your read and a conversation.
Don’t paste an unpublished draft into tools you don’t control to “double-check.”
Don’t quote the percentage at a writer as if it settles the question.

All four modalities