Tried Proofademic’s Plagiarism Checker on a Real Case. Here’s What I Found.
Noticed last week that Proofademic added a plagiarism checker to their AI detector – they’ve been on my radar for the AI detection side and I didn’t expect to see plagiarism checking in there too.
Ran it on a student submission I was already suspicious about – an essay with some sections that felt off but came back clean on our usual tool. Proofademic flagged a 61% match to a source the student hadn’t cited. Not a famous source, not Wikipedia – an obscure blog post from 2023. Our usual tool missed it entirely.
I’m not ready to call it a Turnitin replacement and the UI for the plagiarism report is less detailed than what I’m used to. But as a second pass on cases where something feels off, it caught something real.
Anyone else run it for plagiarism specifically? Curious what the accuracy looks like across different submission types.
5 Replies
Join the discussion.
Log In to Reply
been using it for AI detection for a few months, just started running the plagiarism check this week. the combined report is convenient - one upload, one dashboard. my early impression is that the plagiarism coverage is solid for English-language sources but I haven't tested it seriously on French content yet.
curious about the database size. Turnitin's value is partly the depth of their source index - student paper repository, academic journals, internet content going back years. a newer tool catching obscure blog posts is promising but what's the coverage like for academic sources? that's where the real risk is at the secondary level.
the workflow argument is real. I'm currently running Originality for AI and Copyleaks for plagiarism on anything suspicious. two logins, two reports, two interpretations. a single tool that handles both cleanly would save real time even if neither check is best-in-class.
the sentence-level breakdown in their AI detection is what sold me on Proofademic originally. if the plagiarism report has the same granularity - showing exactly which sentences matched and to what source - that's much more defensible in a parent conversation than a percentage score.
that's a useful data point. the 'catches something the other tool missed' case is exactly what would make me take it seriously. how long did the plagiarism report take to generate? and was the matched source actually visible in the report or just flagged as a percentage?