Can applicant tracking systems detect AI-generated resumes?
Detecting AI writing is not a documented standard feature of applicant tracking systems, and AI text detectors are unreliable. That does not make AI use invisible. Recruiters notice generic phrasing and claims you cannot back up. If you used ChatGPT, make every line specific and true, and you have nothing to hide.
Free, and you see your score without an account.
What the software is built to do
An applicant tracking system stores applications, reads the text of each resume into a profile, and lets recruiters search, filter and rank candidates. Its job is to find people who match the role. Who wrote the words is not part of that job. For the full picture, see what an ATS is.
What vendors actually document
We looked for vendor documentation of a feature that flags resumes as AI-written. We did not find one described as a standard feature. What we did find is about fraud, not authorship:
- Greenhouse documents a fraud detection feature that looks at device and contact signals, such as IP address, browser information, email address and phone number, and gives the employer a fraud report. Its product page describes this as screening "using objective digital signals to help identify potential identity misrepresentation". Neither page describes analysing the text of your resume for AI writing.
That fits what employers worry about: fake applicants, not real people who used a writing tool. Other systems may offer something we could not see, and an employer can always plug in a separate tool. So the accurate answer is "not documented", not "impossible".
Several articles say particular systems now score resumes for AI content. We could not confirm those claims on any vendor's own site, so we leave them out.
Why AI detectors are a weak test anyway
Tools that claim to detect AI writing guess from how predictable the text is. Resumes are short, formulaic and full of standard phrases, which is exactly the kind of text these tools struggle with. In a 2023 Stanford study, seven widely used detectors flagged, on average, 61% of 91 TOEFL essays written by non-native English speakers as AI-generated, while 88 essays by US eighth graders were classified almost perfectly. A careful employer would not reject someone on a detector score alone.
What recruiters do notice
People are the real test. The things that give away a careless AI resume are the same things that weaken any resume:
| What a recruiter sees | Why it hurts |
|---|---|
| "Results-driven professional with a passion for excellence" | Says nothing about you. Could be anyone. |
| Every bullet the same length and shape | Reads as filler, not as a record of real work. |
| Numbers with no story ("increased efficiency by 37%") | They will ask how, and you need an answer. |
| Skills or tools you have not used | Comes out in the first technical question. |
| The job posting pasted back almost word for word | Looks like keyword stuffing, not experience. |
If you used ChatGPT, do this
- Check every fact. Dates, titles, employers, numbers. Delete anything the AI added that did not happen.
- Replace generic lines with specific ones. Swap "managed projects" for what the project was, how big it was and what changed because of you.
- Make it sound like you. Read it aloud. If you would not say a sentence in an interview, rewrite it.
- Keep the job's words where they are true. Matching the posting's language helps recruiters' searches. Inventing experience to match it does not.
- Check the file. Make sure the formatting survived the copy and paste, and that the software can read it. Our checker's full report is free and needs no account.
A resume you can defend line by line is safe whoever typed the first draft.
Questions
Does ATS detect AI?
Not as a standard, documented feature in the vendor documentation we read. An employer can still add a separate tool, and nobody outside the company can rule that out.
Do employers run resumes through AI detectors?
Some might, but it is not documented as common practice, and the detectors themselves misfire. In a 2023 study, seven popular detectors labelled, on average, more than half of 91 essays written by non-native English speakers as AI-generated.
Can recruiters tell if I used ChatGPT?
Often they can tell a resume is generic, which is the real problem. A resume full of vague, polished lines that could describe anyone reads as low effort, whoever wrote it.
Is it wrong to use AI to write my resume?
Using it to draft, tighten or fix grammar is common and fine. Letting it invent experience, numbers or skills is not, and those claims tend to fall apart in an interview.
Check your resume
Free, and you see your score without an account.