About
AI detection has quietly become a decision-making tool. A percentage appears next to a student’s essay and it gets treated as a finding. It is not one — it is an estimate of how closely writing resembles patterns common in machine text, and plenty of human writing sits in that region.
The people who absorb the cost of that confusion are disproportionately the ones who did nothing wrong: second-language writers, students taught rigid essay templates, and anyone who edits their work until it reads cleanly.
What we do differently
- We show which signals produced a score, so you can disagree with the reasoning rather than the number.
- On images we read provenance — signed credentials, embedded generation records — because that is evidence rather than a guess.
- We label vendor accuracy claims as vendor claims, and we leave our own numbers blank until we have measured them.
- We refuse to sell an outcome. No tool can guarantee a detector verdict, and claiming otherwise is how this industry lost its credibility.
What we are not
We are not affiliated with Turnitin, GPTZero, Copyleaks, Originality.AI or any other tool discussed on this site, and we cannot reproduce their scores. Everything on our detector pages comes from public documentation, labelled as such.