NTRVSTA / Stop AI resume spam

Solutions · AI application spam

Stop AI resume spam without rejecting real candidates

Auto-applied, AI-written applications now bury the real candidates in many reqs. Here is a four-layer way to cut the spam: lock down your ATS, score every applicant against your scorecard with written reasons, make the first round a live conversation with follow-up questions, and verify identity before the offer.

Last updated November 21, 2026.

What changed: more applications, less signal

Applying got cheap. Candidates can now generate a tailored resume and cover letter in seconds, and auto-apply tools send them to hundreds of postings. The volume lands on recruiters who still review applications one at a time.

  • Greenhouse reported its customers received an average of 222 applications per job opening in Q1 2024, almost three times the level at the end of 2021.
  • In a Greenhouse survey published November 19, 2025, 34% of recruiters said they spend up to half their week filtering spam and junk applications, and 65% of hiring managers said they had caught applicants using AI deceptively.
  • LinkedIn found 73% of HR professionals say less than half of the applications they receive meet all of the job criteria.
  • Gartner press release reported that 39% of candidates in its survey used AI during the application process, and it predicts that by 2028 one in four candidate profiles worldwide will be fake.

Most of this volume is not fraud. It is real people applying to jobs they are not a fit for, at scale. A smaller share is fabricated experience or a different person showing up to the interview. The fix is different for each, which is why one filter is never enough.

A four-layer defense against AI resume spam

  1. Close the door on bots and repeat floods (your ATS)Turn on the job-board spam protection, application limits, duplicate checks and knockout questions your ATS already has. These are free, fast and they act before a recruiter sees anything. Our ATS playbooks below show the exact settings.
  2. Score every applicant against your scorecard (Carolina)Carolina, NTRVSTA's AI recruiting agent, reads every applicant against the must-haves, nice-to-haves and knockouts you set, writes down the reasons for each score, and ranks the pool. Keyword-stuffed resumes stop jumping the queue because the reasons have to hold up, and your team starts at the top of the list instead of at the top of the inbox.
  3. Make the first round a real conversation (Carolina)Carolina runs live first-round interviews by video or voice, at any hour. Each question adapts to the previous answer, and strong answers get harder follow-ups. A memorized or relayed answer is much harder to sustain three follow-ups deep, and the transcript and scorecard land in your ATS for a person to review.
  4. Verify identity at the step where it matters (your team)Use an ID check from your background-check provider before an offer, and keep at least one live, on-camera conversation with a person in the process. The FBI has warned since 2022 about deepfakes and stolen identities in remote job applications.

What Carolina does and does not do

Carolina is built to cut the hours between "applied" and "worth your time". She is not an identity or fraud product, and we do not claim she detects deepfakes or proxy interviewers. Here is the honest split:

Carolina doesScores every applicant and sourced profile against your scorecard with written reasons; runs live first-round interviews with adaptive follow-up questions; writes scorecards and transcripts to your ATS; messages and reminds candidates by email, SMS or WhatsApp; posts a morning digest in Slack or Teams.
You controlEvery action can be set to suggest, ask first or run on its own. Offers and comp changes always wait for a person.
Pair it withYour ATS's own spam and fraud settings, and an identity check from a background-check provider. Carolina connects to Checkr, Sterling and HireRight.

ATS playbooks: cut AI-generated applications at the source

Each guide uses only features documented in the vendor's own help center, with links.

Worried about the interview step itself? Read how to detect proxy candidates in remote interviews.

What AI application spam costs you

Put your own numbers into the AI application spam cost calculator: open reqs, applications per req, the share you judge low-fit or fake, and review time. It shows recruiter hours and dollars per month, with every assumption labeled.

Where NTRVSTA is not the right fit

If your problem is identity fraud on a small number of high-risk roles, start with an identity-verification vendor and your ATS's fraud settings; Carolina is not that tool. If you only get a few dozen applications per req and your recruiters keep up, ATS knockout questions may be all you need.

Questions

How do I stop AI-generated resumes from flooding my reqs?

Layer the fixes. Turn on your ATS spam protection, application limits and knockout questions first, then score every remaining applicant against a written scorecard, and make the first round a live conversation with follow-up questions. No single filter catches both mass-applied resumes and fabricated experience.

Can AI detect a fake or AI-written resume?

Some ATS vendors offer fraud signals such as device, IP and email checks, and treat them as flags for a person to review, not verdicts. Text alone is a weak signal because many honest candidates use AI to polish a resume. Scoring against evidence and a live follow-up conversation tell you more than guessing who wrote the words.

Does Carolina detect proxy candidates or deepfakes?

No, and we do not claim it does. Carolina's live interviews ask adaptive follow-up questions, which makes memorized or relayed answers harder to sustain and gives a reviewer a transcript to check. Pair that with an identity check from your background-check provider and a live conversation with a person before an offer.

Will stricter filters reject good candidates?

They can. Keep knockout questions to true requirements, review a sample of auto-rejected applications each month, and make sure every rejection has a reason you could explain. Colorado, Illinois, NYC and California also have rules when AI influences hiring decisions; see our compliance hub.

Sources

‍