Statistics & Trends
Artificial intelligence has lowered the barrier to fraud in nearly every domain — and hiring is no exception. What was once a niche threat has become a documented, growing problem affecting organizations of every size. This post compiles the key statistics and trends on AI-powered job fraud so hiring teams can understand the scale of what they're up against.
The Numbers
In 2018, generating a convincing real-time deepfake required significant computing resources and technical expertise. By 2023, consumer-grade tools available for under $20/month could produce real-time face-swapping capable of passing a casual video call. The democratization of this technology has put it within reach of fraudsters with minimal technical skill.
The widespread shift to remote hiring removed the step where candidates came to an office, presented a physical ID, and were seen in person. Without that checkpoint, the video interview became the primary — and often only — opportunity to verify identity. That's a single point of failure in a process that carries serious financial and security implications.
AI isn't just being used to fake faces — it's being used to generate entirely synthetic professional identities. Fabricated LinkedIn profiles, AI-written portfolios, and constructed reference networks can make a completely fictitious candidate appear credible enough to reach the interview stage. By the time a deepfake face is needed, the rest of the fraudulent identity is already in place.
Organizations that hire exclusively through video have no in-person checkpoint. Software engineering, IT security, and DevOps roles are disproportionately targeted because they offer high salaries, remote work, and immediate access to sensitive systems. The FBI's public advisory specifically called out IT and software roles.
Agencies that source candidates for client organizations face compounded risk: a single fraudulent placement affects both the agency's reputation and the client's security. The North Korea case — documented by the DOJ in 2024 — involved staffing intermediaries who had no verification step at placement.
Organizations operating under security clearance requirements or financial regulations face regulatory consequences — not just operational ones — when identity verification fails.
Every indicator points in the same direction. Deepfake generation quality is improving. The tools are getting cheaper. Remote hiring is not declining. And the fraud techniques are evolving faster than most organizations' awareness of them. The lag between when a new method appears and when it becomes widely understood is precisely the window in which the most damage is done.
The 0.1% human detection rate isn't a training problem — it reflects a fundamental limitation of human visual perception against AI-generated content. The solution has to be systematic and applied before the interview, not during it. Verification needs to happen before a candidate ever receives your meeting link — not after you've invested hours evaluating someone you can't confirm is real.
Don't rely on your eyes. Use the data.
Stop Deepfake Candidates verifies candidate identity before the interview using government ID matching and deepfake detection — not human judgment. $5 per verification, no subscription.
Start Verifying Free →