Catching fake profiles with the Fraud Detector

Last updated: August 16, 2026

Fraud Detector is an AI teammate that screens candidates for fake credentials, fraudulent patterns, and synthetic identities before you spend time reviewing or interviewing them. It analyses resumes, email addresses, phone numbers, social profiles, and network signals to flag likely fraud.

Who can do this: Admin and Super Admin enable it and add it to a job. Recruiter and Hiring Manager can run manual checks, view results, and change a fraud status.

Where to find it: Settings → AI Teammates → Fraud detector (turn it on for your organisation), then Jobs → a job → Settings → AI Teammates tab (add it to that job).

Turn it on

Fraud.png

  1. Go to Settings → Kula Teammates → Fraud detector and toggle on Enable fraud detector.

  2. Open the job you want it on, go to Settings → AI Teammates tab, and click Add teammate on the Fraud detector card.

  3. Choose Manual or Automatic mode. Automatic runs on every eligible inbound candidate; Manual only runs when you trigger it.

  4. Optionally turn on Auto-reject for candidates marked Suspicious.

  5. Click Save.

Turning this on shows applicants an AI usage disclaimer on your career page.

What it needs to run

3.png

A fraud check needs a resume, email address, and phone number on the candidate's record. If any of those is missing, the check can't run — this is also why Automatic mode won't cover every applicant: it only runs on candidates who already have all three on file.

Run a check

  • One candidate: open their profile, click the three-dot menu, select Review with Fraud detector, and click the fraud status badge once it finishes to see the results.

    1.png

  • Several at once: in the Candidates tab, select candidates, open the bulk action menu, and select Review with Fraud detector — a progress widget shows you how it's going.

    2.png

  • Re-run a check: if the previous evaluation Failed, or if the candidate's resume, email, phone number, education, or work experience has changed since the last check.

Understand the statuses

  • Not evaluated: No check has run for this candidate yet.

  • In progress: The check is currently running.

  • No issues: No significant fraud signals found.

  • Inconclusive: Manual review recommended — the check didn't reach a confident answer either way.

  • Suspicious: High-confidence fraud indicators found.

  • Failed: The evaluation didn't complete (often missing prerequisites) — no credits are consumed.

Click a candidate's fraud status badge to open the Fraud Detection panel: overall evaluation, signal categories, risk levels, explanations, an audit trail, and the previous evaluation if there was one.

Override a result

Open the Fraud Detection panel, click Change fraud status, enter a reason (required), and save.

Auto-reject

4.png

If Auto-reject is on, a candidate marked Suspicious is rejected automatically with the reason "Fraudulent candidate." Recruiters can reactivate the candidate or edit the rejection afterward if it turns out to be a false positive.

Auto-reject only fires on a candidate's first automatic pass — a later manual review, or a recruiter manually changing someone's status to Suspicious, won't trigger it again.

Turn it off for a job

Click the delete icon on the Fraud detector card and confirm Disable. Existing fraud statuses on candidates already checked stay as they are.

Good to know

  • Each check costs 3 credits and takes about 30–50 seconds. Failed checks don't consume credits.

  • There's no self-serve view of your remaining credit balance. Admins can request more credits; non-admins should notify an admin if the team is running low.

    5.png

  • Fraud status is available as a filter, group-by, and breakdown option in Current job pipeline, Pipeline history, and Contacts reports.

FAQ

  • Why is it that I set the fraud detector to automatic but it doesn't seem to run on all applicants — are there parameters?

    Automatic mode only runs on candidates who already have a resume, email, and phone number on file — those three are required for a check to run at all, so anyone missing one of them won't get an automatic check until that information is added.

  • Do you know yet how many fraud credits I have left?

    Not from the product today — there's no in-app view of your remaining credit balance. Admins can request more credits directly; if you're not an admin, ask one on your team to check or top up.

Need help?

If you have questions or need assistance, reach out to us at support@kula.ai or use the in-app chat.