Guide
Fake candidates: how to spot them, and how to stop meeting them
Facts checked 2 November 2026
A fake candidate is an applicant who isn't who, or what, they claim to be: a resume with invented experience, a stand-in on the interview call, a deepfake face, or a stolen identity used by an organised scheme such as North Korea's remote IT workers. Gartner expects one candidate profile in four worldwide to be fake by 2028,1 and 91% of US recruiters have already spotted candidate deception.2 Below: what counts as fraud, the numbers, the warning signs, how to check without insulting real applicants, and why going out to people with a real work history avoids most of it.
Where Boost fits
Keep the inbox shut: HireAll Boost
Fraud gets in through applications. Boost doesn't take any: it picks whom to invite from public work histories, against your must-haves, and you meet only people who said yes and answered your questionnaire.
- Nobody gets into your pipeline by applying
- Each person checked against your must-haves before the invite
- Your questionnaire answered before any call
What counts
What counts as a fake candidate
Candidate fraud runs from a polished lie to an organised crime. Using AI to write a resume isn't fraud on its own: 39% of candidates in a Gartner survey used AI while applying, most often to write resume text.1 It becomes fraud when the experience, the person or the identity behind the application is false.
Invented experience
A resume, often AI-written, that claims jobs, titles or skills the person never had, tuned to the keywords of the job post.
Hidden instructions
Text a person never sees, such as white type, written to steer an AI screener. 22% of US hiring managers have caught it in resumes.2
Proxy interviews
Someone else takes the interview, or the candidate takes it for someone else. 6% of candidates in a Gartner survey admitted to it.1
Deepfake interviews
A face or a voice altered in real time on the call. 18% of US hiring managers have caught candidates appearing as deepfakes.2
Stolen and rented identities
Real people's names, addresses and documents, bought, stolen or rented, so the background check passes.4
State-run remote workers
North Korean IT workers who get hired under false identities; the FBI says the regime takes an estimated 85–90% of what they earn.4
The numbers
Candidate fraud in numbers
Surveys, a forecast and court cases, each with its source. The surveys ask people what they did or saw, so they measure what gets noticed, not everything that happens.
| Figure | What it measures | Who says so |
|---|---|---|
| 1 in 4 | Candidate profiles worldwide that will be fake by 2028: a forecast | Gartner, July 20251 |
| 6% | Candidates who admitted to interview fraud, posing as someone else or letting someone pose as them | Gartner survey of 3,000 candidates1 |
| 91% | US recruiters who have spotted candidate deception | Greenhouse, November 20252 |
| 65% | US hiring managers who caught applicants using AI deceptively: reading AI scripts 32%, hidden prompts 22%, deepfakes 18% | Greenhouse2 |
| 34% | Recruiters who spend up to half their week filtering spam and junk applications | Greenhouse2 |
| 11,000 | Applications a minute on LinkedIn, up more than 45% in a year | The New York Times, June 20255 |
| 291 | Applications per hire, against roughly 100 in early 2021 | Ashby6 |
| 70+ | US employers hit by one North Korean remote IT worker scheme | US Justice Department, October 20263 |
Volume is what makes fraud pay. When a job post draws hundreds of applications, nobody reads each one closely, and a fake that looks like the job description gets through the first filter. Fraud rings know it: the FBI describes North Korean workers paying freelancers to apply "to as many open positions as possible".4
Warning signs
Warning signs of a fake candidate
None of these proves anything alone, and most have innocent explanations: the FBI warns that a single indicator can't identify a fraudster and that aggressive filters turn away qualified people.4 Look for several together, then check.
A resume that mirrors the post
Every keyword of the job description, in its order, and little you could ask a follow-up question about. Ask about one project in depth: invented experience runs out of detail fast.
A history that doesn't hold up
Dates, titles or employers that differ between the resume, the profile and what the person says. Claiming someone else's professional profile as one's own is a known tactic.4
Answers read from a script
Pauses before each answer, eyes moving across a screen, fluent generalities that fall apart at the first "why". 32% of US hiring managers have caught candidates reading AI-written answers.2
Camera trouble on cue
Video off, frozen or out of sync with the voice, a face that blurs when the person turns. 36% of US job seekers say they've altered their appearance, voice or background in a video interview, not always to deceive.2
A different person each round
A voice, a face or a level of skill that changes between interviews. Fraud schemes pay real people to appear on camera and sit interviews for them.4
Reluctance to verify
Excuses to avoid an in-person meeting, a live ID check or official documents. The FBI names E-Verify, a US passport, several forms of ID and live verification as checks that raise friction for fraudsters.4
Logistics that keep changing
A laptop to be shipped to a different address, bank details or addresses that change often. Schemes run "laptop farms" and rotate accounts after bans.4
After the hire: remote control
Remote desktop software or KVM-over-IP devices on a work laptop, and apps that have nothing to do with the job.4
How to check
How to check a candidate without insulting the real ones
The aim is friction for fraud, not suspicion of everyone. Gartner recommends saying up front how you check, and found that 62% of candidates are more likely to apply when an employer requires in-person interviews.1 The FBI calls the approach "identity in depth": several layers of checks through the whole hiring and employment cycle, rather than one filter at the door.4
- Say how you hire. Tell candidates early which checks you run and that you run them for everyone. Real applicants don't mind; fraudsters drop out.
- Ask about the work, not the resume. One project, in depth: what went wrong, what they'd change, who else was on it. Follow-up questions are hard to fake live.
- Keep the same questions across rounds. If the second interviewer hears a different story, or a different voice, you'll know.
- Verify identity live before the offer. A live ID check, documents and, in the US, E-Verify.4
- Meet once in person if you can. Even one on-site step filters out most stand-ins.1
- Ship equipment only to a verified address, and keep an eye on remote-access tools after the start.4
Prevention
Why outbound sourcing avoids most of it
Every scheme above enters the same way: through an application. The inbox is open to anyone, LinkedIn alone takes about 11,000 applications a minute,5 and a fake that matches the job post costs a fraudster nothing to send a thousand times.
Outbound sourcing turns that around. Instead of reading whoever applied, you choose whom to contact, from people with years of visible work history who weren't applying anywhere. Nobody gets into your pipeline by flooding it, and the person you contact is the one whose history you read. A short questionnaire before the first call then shows you real answers before you spend an hour on video.
It isn't airtight, and it's worth saying so. The FBI reports schemes renting real people's professional accounts, with years of genuine activity, to apply for jobs.4 An aged profile is a good sign, not proof, so keep the live ID check before the offer whichever way a candidate came in.
This is how HireAll Boost works. It reads profiles in full against the must-haves you set and invites only the people who meet them; nobody can apply to a Boost role. The people who say yes answer your questionnaire, then book a slot in your calendar, so you've read their answers before you meet. You start with 100 free credits, one credit per person invited.
FAQ
Fake candidates: questions people ask
What is a fake candidate?
An applicant who isn't who, or what, they claim: invented experience on a resume, someone else in the interview, a deepfake on the call, or a stolen identity. At the far end are organised schemes such as North Korea's remote IT workers.4
How common is candidate fraud?
Common enough that 91% of US recruiters have spotted candidate deception2 and 6% of candidates admit to interview fraud.1 Gartner expects one candidate profile in four worldwide to be fake by 2028; that is a forecast, not a measurement.
How can you tell if a resume is AI-generated?
Often you can't, and it doesn't matter much: many honest candidates use AI to write.1 What matters is whether the experience is real. Ask about one project in depth, and check the dates and employers against the person's public profile and references.
What is a deepfake job interview?
An interview in which the candidate's face or voice is altered in real time, to look like someone else or to hide who is speaking. 18% of US hiring managers in Greenhouse's survey have caught one.2 An in-person step, or a live ID check on camera, is the usual counter.
Who are the North Korean IT workers?
Remote software and IT workers, controlled by the North Korean state, who get hired by Western companies under stolen or rented identities, with help from paid facilitators; the regime takes an estimated 85–90% of their earnings.4 One scheme charged in October 2026 reached more than 70 US employers.3
How do I verify a remote candidate's identity?
In layers: a live ID check on camera, official documents and, in the US, E-Verify before the offer; consistent questions across interview rounds; equipment shipped only to a verified address.4
Does outbound sourcing stop fake candidates?
Most of them, because nobody gets in by applying and you choose whom to contact from their work history. Not all: fraudsters sometimes rent real profiles with years of history,4 so keep a live ID check before the offer.
Is using AI to write a resume fraud?
No. 39% of candidates use AI while applying, mostly to write resume and cover letter text.1 It becomes fraud when the experience it describes is invented, or when the person in the interview isn't the one who applied.
Read next
Stop sifting. Start hunting.
Describe the role, pick a pace and meet the first people who said yes.
Get 100 free credits →No card. 1 credit = one candidate who gets an invite.
Sources
- Gartner, press release, 31 July 2025: "by 2028, one in four candidate profiles worldwide will be fake". A forecast. In its surveys 6% of candidates admitted to interview fraud, "either posing as someone else or having someone else pose as them"; 39% used AI while applying, 54% of those to write resume text; 62% are more likely to apply where interviews are in person. Checked 2 November 2026
- Greenhouse, press release, 19 November 2025: 34% of recruiters spend up to half their week filtering spam and junk applications. In the US, 91% of recruiters have spotted candidate deception; 65% of hiring managers caught applicants using AI deceptively: reading AI-generated scripts 32%, prompt injections in resumes 22%, deepfakes 18%; 36% of job seekers altered their appearance, voice or background in video interviews. Survey of 4,136 people in the US, UK, Ireland and Germany. Checked 2 November 2026
- US Department of Justice, press release, 13 October 2026: one of the North Korean remote IT worker schemes charged "impacted more than 70 U.S. company employers". Checked 1 November 2026
- FBI, Internet Crime Complaint Center, North Korean IT Workers Expanding Use of Facilitators, Broadening Recruitment, 13 October 2026: IT workers "cannot be identified reliably with a single indicator"; the regime "receives an estimated 85-90% of the revenue"; they pay Americans to "appear on camera, participate in job interviews", rent facilitators' professional networking accounts "which often contain years of legitimate activity history", pay freelancers "to apply manually to as many open positions as possible", run laptop farms and "rotate through bank accounts"; E-Verify, a US passport, several forms of ID or live verification "can raise friction"; aggressive controls risk "false positives". Checked 2 November 2026
- The New York Times, DealBook, Employers Are Buried in A.I.-Generated Résumés, 21 June 2025: applications on LinkedIn have "surged more than 45 percent in the past year"; the platform is "clocking an average of 11,000" a minute. Checked 2 November 2026
- Ashby, Recruiter Productivity Trends, reported by HR Dive, May 2026: "291 applications per hire, compared to roughly 100 in early 2021". Checked 1 November 2026