An AI detector is a tool that claims to tell whether a piece of text was written by AI. They are unreliable: they miss AI text that has been lightly edited and wrongly flag human writing, especially from non native English writers. Do not base hiring decisions on them.
AItopic
2 minread
1related terms
In one line
A tool that claims to tell whether text was written by AI. They are unreliable and wrongly flag non native English writers, so do not base hiring decisions on them.
Most detectors look at how predictable the words are. AI text tends to use common word choices, so very smooth, simple writing scores as likely AI. The problem is that many people also write that way, including people writing in a second language or following a template. A Stanford study found popular detectors misclassified many essays by non native English writers as AI generated while doing much better on native writers' work.
In recruiting, detectors are tempting because AI written resumes and cover letters are now common. But a resume written with AI help can still describe real experience, and a fully human resume can be flagged by mistake. Rejecting people on a detector score risks losing good candidates and treating some groups unfairly. The better question is whether the content is true and relevant, not who typed it.
Why it matters to recruiters
Recruiters face more polished, AI assisted applications than ever. Using detectors to filter them looks efficient but creates false rejections and fairness risks, since errors fall harder on some groups. Knowing their limits pushes teams toward checks that work: structured screening questions, skills assessments and interviews that test what a candidate really knows.
An example from a recruiting desk
A hiring manager runs cover letters through a detector and wants to reject anyone scoring over 70 percent AI. The recruiter checks three flagged letters and finds two from candidates who wrote in their second language. In this example the team drops the detector and adds two short, job specific screening questions answered live on the phone screen instead.
Tips that help
Do not use detector scores to reject or rank candidates.
Test skills directly with work samples, assessments or structured interview questions.
Ask candidates to talk through items on their resume to confirm the experience is real.
Set a clear, fair policy on AI use in applications and share it with candidates.
Mistakes to avoid
Treating a detector's percentage as proof that a candidate cheated.
Penalizing well written applications when the experience they describe is genuine.
Frequently asked questions
Can you tell if a resume was written by AI?
Not reliably. Detectors make frequent mistakes, and AI generated text that has been edited is hard to spot. Signs such as generic wording or perfect formatting are weak clues at best. Focus instead on whether the experience is real and relevant, which you can test in a screen or interview.
Is it fair to reject candidates for using AI on a resume?
Usually not, unless your stated rules forbid it for a good reason. Many candidates use AI the way they once used templates or a friend's help. What matters is accuracy. Rejecting people based on detector scores also risks unfair treatment, since detectors misjudge some groups more than others.
What should recruiters do instead of using AI detectors?
Ask job specific questions that need real knowledge to answer, use short skills tests where they fit the role, and verify key claims through references or background checks. A live conversation about past work quickly shows whether someone understands what is on their resume.
