Gartner predicts one in four candidate profiles will be fake by 2028 — and 91% of recruiters say they have already spotted deception. Here is the verification playbook for hiring real people in the age of synthetic applicants.
A few years ago, interview fraud sounded like an edge case — the occasional embellished CV, the rare candidate who had a friend sit a phone screen. In 2026, it is an industry. Gartner now predicts that by 2028, one in four candidate profiles worldwide will be fake, and in its survey of 3,000 job seekers, 6% openly admitted to interview fraud — impersonating someone else or having someone interview on their behalf. Given that not everyone confesses to cheating in a survey, treat that figure as a floor.
For technology hiring — remote-first by default, fast-moving, and built around scarce, well-paid skills — the problem is most acute of all. This is the story of how AI turned candidate verification into the defining operational challenge of tech recruitment, and what serious employers (and honest candidates) should do about it.
How big is the problem, really?
Bigger than most hiring teams assume, and growing faster. Deepfake fraud attempts in hiring jumped by roughly 1,300% between 2023 and 2024, and the trajectory has continued to accelerate since. In Greenhouse’s 2026 AI Hiring Report, 91% of recruiters and hiring managers said they had spotted or suspected candidate deception, and 74% said they were more worried about fake credentials than a year earlier. A separate Greenhouse survey of more than 4,000 hiring managers found that 31% had personally interviewed a candidate they suspected — or confirmed — was using a fake identity via deepfake technology.
When the security firm Pindrop posted a single job listing as an experiment, roughly 12% of all applicants — about 100 of 827 — used fake identities. And the threat is not limited to opportunists. The US Department of Justice has pursued coordinated actions against North Korean remote-IT-worker schemes involving stolen identities and “laptop farms”, alleging that more than 300 US companies unknowingly hired impostors for remote roles. The consequences of a successful fake hire run well beyond a salary paid for no work: intellectual property theft, data exfiltration and malware installed from inside the perimeter.
The taxonomy of AI-enabled deception
It helps to think of candidate fraud as a spectrum rather than a single behaviour, because the response should differ at each level.
- AI-polished applications. Somewhere between 40% and 80% of applicants now use generative AI to draft CVs, cover letters and application answers. At its mildest this is formatting help; at its worst it is systematic exaggeration — the most common deception recruiters report is AI-generated CV embellishment.
- Live interview assistance. Real-time AI coaching tools feed candidates answers during video interviews. Combined with auto-apply services that fire off thousands of applications, this floods pipelines with candidates whose on-paper and on-screen performance bears little relation to their actual ability.
- Proxy interviewing. A stronger engineer sits the technical interview; a different person turns up for the job. Fraud-on-demand services now sell exactly this.
- Synthetic identities and deepfakes. The far end of the spectrum: AI-generated faces and cloned voices layered over a real operator, with fabricated work histories and stolen or invented identity documents. As Pindrop’s CEO put it, all that is needed today is a static image and a few seconds of audio.
Why “ban AI” is the wrong answer
The instinctive employer response — forbid candidates from using AI at all — fails on two counts. First, it is unenforceable. Second, it is inconsistent: candidates are largely responding to an environment employers created. When companies screen thousands of CVs with algorithms, job seekers rationally optimise their materials for those algorithms. Penalising candidates for AI use while your own funnel is AI-driven reads as hypocrisy, and it deters exactly the strategic, tool-literate professionals that AI and data teams want to hire.
The productive question is not whether a candidate used AI, but whether their underlying skills, experience and identity are genuine. That reframing points to verification, not prohibition.
The verification playbook
No single control stops modern candidate fraud; what works is a layered defence across the hiring lifecycle. The strongest programmes we see in European and Asian tech hiring combine five elements.
1. Set expectations publicly
State in job adverts and interview invitations what AI use is acceptable (research, preparation, polishing) and what is not (live assistance, impersonation), and say plainly that identity verification and fraud detection are part of the process. Gartner’s research suggests transparency itself deters: 62% of candidates say they are more likely to apply to a role that includes in-person interviews — honest candidates read verification as a signal of seriousness, not suspicion.
2. Verify identity early, professionally
Government-ID verification at the pre-screen stage, handled by a proper verification platform and framed as standard onboarding rather than accusation, removes the bulk of synthetic identities before they cost interview hours. Consistency checks matter too: does the LinkedIn history, the CV, the email domain and the stated location line up?
3. Design interviews that filters cannot survive
Low-grade deepfake filters break under simple, polite requests: turn the camera slightly, adjust the lighting, share a screen mid-conversation. More fundamentally, structured live problem-solving — working through an unfamiliar, role-realistic scenario with follow-up questions that probe the reasoning — is very hard to fake, whether the assistance is a proxy or a chatbot. Several major employers, including Google and McKinsey, have reintroduced mandatory in-person rounds for precisely this reason; for remote-first hiring, a supervised live working session is the practical equivalent.
4. Triangulate references
AI-generated references are among the fastest-growing fraud vectors — nearly half of recruiters in Greenhouse’s research reported encountering fake references. Verify referees independently: find them through the company’s official channels or LinkedIn rather than solely through details the candidate supplies, and speak to them by video where stakes are high.
5. Keep validating after the offer
Fraud that survives hiring should still be caught at onboarding: background checks, right-to-work verification, device and access monitoring, and anomaly alerts in the first weeks. Security teams increasingly treat a new remote hire as an identity-assurance event, not just an HR one.
The specialist recruiter as a human verification layer
There is a reason demand for specialist agencies has risen alongside candidate fraud. A recruiter who works a narrow market — machine learning, data engineering, MLOps — knows the community: who actually worked on which team, which claims ring true, which referee is real. At Acquism we interview candidates on video before any client does, verify identity and work history as standard, and test technical claims against consultants who understand the technology. In a market where a quarter of profiles may soon be synthetic, a pre-verified shortlist is not a luxury; it is risk management.
A note for honest candidates
If you are a genuine job seeker, none of this should worry you — but it should shape how you use AI. Using AI to research a company, structure your CV or rehearse answers is legitimate and increasingly expected. Letting AI answer live interview questions, inflating titles and metrics, or presenting generated work as your own is career-damaging: verification is improving faster than fraud, employers share intelligence, and specialist recruiting communities are small. The candidates who win in 2026 are those who use AI openly as a productivity tool and let their real, verifiable skills carry the interview.
Hiring for AI, data and engineering roles and want a verified shortlist rather than a pipeline of question marks? Acquism’s consultants screen, verify and technically assess every candidate we represent, across permanent, contract and executive search mandates in Europe and Asia. Get in touch to talk about your next hire.