Hiring processRead at source, 24 September 2026

Fake Job Applicants: How to Spot and Screen Them Out (2026)

A fake job applicant is an application that cannot turn into a hire: a bot or script submission, a duplicate, a person using someone else’s identity, or a real person whose contact details, location or credentials are wrong. In hourly hiring the expensive version is rarely a deepfake. It is the steady share of junk, duplicate and unreachable applications that eats recruiter time and inflates your real cost per applicant. Gartner found that 6% of 3,000 candidates it surveyed in 2025 admitted to interview fraud, and it predicts that by 2028 one in four candidate profiles worldwide will be fake, but that risk concentrates in remote, credential-heavy roles. For frontline jobs, one or two knockout questions, a text within minutes, a same-day phone call and an in-person interview remove most of it without adding friction for the real applicants.

What fake applicants look like in hourly jobs

"Fake" covers several different problems, and each one is caught at a different point. Treating them as one problem is how employers end up adding a CAPTCHA and a résumé upload that stop almost none of them and turn away real applicants.

The first four are the ones frontline employers see week to week, and none of them involves anyone trying to deceive you. They are what happens when many people apply from a phone in a few seconds. That matters, because the fix for them is follow-up, not suspicion.

Bot and duplicate applications from paid social ads

Where the application lives decides which kind of junk you get. An open form on a careers page or a job board that anyone can submit without signing in is where bots and scripted submissions show up, because there is nothing to stop a script from filling it in a thousand times. A lead form inside Facebook or Instagram is submitted from a logged-in account and is usually pre-filled from that account’s profile, so scripted submissions are much less of a problem there. What paid social ads produce instead is the ordinary list: duplicates from people who saw the ad twice, and contact details that were pre-filled from a profile and are out of date.

Three settings do most of the work on a paid job ad without making the application longer:

• State the requirement in the ad. "CDL-A required," "must be available weekends," "in [City]." Many out-of-area and ineligible applications come from an ad that did not say what the job needs.
• Ask for confirmation, not more fields. Meta’s instant forms offer a higher-intent form type that adds a review screen before the person submits, which is the cheapest way to make someone check a pre-filled phone number. It lowers volume somewhat, so test it rather than assume it.
• De-duplicate on the way in. Match on phone number and email when leads land in your ATS or inbox, and keep the earliest record. Counting duplicates as applicants overstates what the ad delivered.

Deepfake and proxy interview candidates

The alarming numbers in the news are about this category. In a 2025 survey of 3,000 job candidates, Gartner found that 6% admitted to taking part in interview fraud, either posing as someone else or having someone else pose as them, and it warned that candidate fraud creates cybersecurity risks that can be far more serious than making a bad hire. That risk sits mainly in remote jobs, where the whole process happens on a screen and the hire is given system access without ever being met.

Most frontline jobs close that gap by their nature: the person has to turn up to a warehouse, a nursing home or a job site. The hourly version of the risk is narrower. It is the interview that happened by phone with one person and the first day that is attended by another, or a claimed license that was never checked. Two habits cover it. Hold the final interview in person, or on video with the camera on. And complete identity and employment-eligibility paperwork the way the law requires, at hire and not before the offer, checking that the person in front of you is the person you interviewed. Gartner’s own research found that 62% of candidates said they were more likely to apply where the employer requires in-person interviews, so requiring one need not cost you applicants.

Screening them out: knockout questions, text verification, phone callback

The screen below removes most kinds of fake applicant in the table and takes a real applicant no extra time. It works because a real applicant who wants the job answers a text, and an address that is fake, mistyped or uninterested does not.

• Step 1: One or two knockout questions on the form. Ask only about the thing the job legally or physically requires: the license, the shift, the location. Every other question costs applicants. Our guide to pre-screening questions sets out which questions are worth asking and which ones create legal exposure.
• Step 2: A text within minutes. Send an automated text that names your company and the job and asks one question the person has to answer, such as the best time to call. A number that fails to deliver has identified itself. A reply confirms a reachable person who remembers applying. In frontline hiring it is an effective fraud filter, and it is also a step that recovers real applicants who would otherwise go cold. See how it works in text recruiting software.
• Step 3: A phone call the same day. Eight to ten minutes, the knockout facts confirmed out loud, the interview booked on the call. A proxy or scripted applicant rarely survives a live conversation about the shift and the commute.
• Step 4: An in-person interview, or camera-on video for roles that cannot meet in person. Check the name against the application. For credentialed roles, verify the license against the state registry before the offer is final.

What not to do: add friction for everyone to stop a few. A CAPTCHA, an account-creation step or a mandatory résumé upload stops some scripts on an open career-site form, but in our 2026 benchmark of 891 campaigns, what happens after the click explained 70% of the difference in cost per applicant between campaigns. Every step you add to stop a fake applicant is paid for in real ones.

What they cost you

Fake applicants cost you twice: once in the recruiter hours spent calling numbers that go nowhere, and once in a cost per applicant that looks better than it is. The correction is simple arithmetic. Divide your cost per applicant by the share of applicants who turn out to be real and reachable.

$13.88 is the median campaign’s cost per applicant in Boostpoint’s 2026 Social Job Advertising Benchmark. The invalid shares are illustrations, not measurements; put your own in. Measure it by counting, for one month, the applicants whose first text failed or who never answered a text or a call, and dividing by the total. That share is also the most honest way to compare two channels, because a channel with a lower cost per applicant and a higher invalid share can be the more expensive one.

For remote roles the cost can be far larger than wasted hours, because a fraudulent hire may be given access to systems and data. That is the case Gartner’s warning is about, and it is why the identity check at onboarding matters more there than anywhere else.

Sources

Gartner, "Gartner Survey Shows Just 26% of Job Applicants Trust AI Will Fairly Evaluate Them," press release, July 31, 2025 (the 6% interview-fraud figure is from a 2Q25 survey of 3,000 job candidates; the release is also the source of the 2028 prediction, the 62% in-person-interview finding (2Q25) and the cybersecurity warning). Boostpoint 2026 Social Job Advertising Benchmark: 891 Boostpoint-managed campaigns on Meta (1,334 campaign-months); costs are what advertisers paid, inclusive of campaign management; cost per applicant is not cost per hire. Read September 24, 2026.

Frequently asked questions