The new face of candidate fraud and why your reference check is your last line of defence

Hiring has always carried risk. But the fraud problem facing HR and talent acquisition teams today looks nothing like it did five years ago. Artificial intelligence has made it easier than ever for candidates to fabricate their credentials, replicate someone else's voice in a video interview or invent a referee who doesn't exist. And most hiring processes aren't built to catch it.

If your screening process relies heavily on CVs and one-way video interviews, it's worth asking: how confident are you that what you're seeing is real?

What does candidate fraud look like in the age of AI?

The days of a slightly embellished job title or an exaggerated skill set are far from gone, but AI has raised the stakes considerably. Today, candidate fraud can take several forms that are genuinely difficult to detect with the naked eye.

Fake or AI-generated CVs are now trivially easy to produce. Large language models can generate polished, keyword-optimised resumes in minutes, complete with plausible-sounding employers, roles and tenure. These documents can sail through ATS filters and impress recruiters who have no way to cross-reference them in real time.

Deepfake video interviews represent a more unsettling development. Candidates are using AI tools to alter their appearance or voice during live and recorded video interviews, allowing someone other than the applicant to effectively sit the interview on their behalf. In a one-way video format, there's often nothing stopping this.

Fabricated referees are also on the rise. A candidate can set up a false email address, a temporary phone number or even use another AI tool to respond to reference check questionnaires as a fake employer.

A Citi Institute report on AI and deepfakes estimates that as many as one in four candidate profiles globally could be fake by 2028. That's not a niche concern – it's a systemic one.

How widespread is candidate fraud in hiring today?

The data is hard to ignore. According to a survey by TechRadar, only 19% of hiring managers feel confident they can spot a fraudulent candidate. That means the vast majority of hiring teams are operating without the tools, training or processes to reliably detect fraud when it occurs.

Separately, a Gartner survey of 3,000 candidates found that 6% admitted to some form of interview fraud. That figure represents candidates who acknowledged it. The true number is almost certainly higher.

‘What we're seeing across our platform is a clear shift in the nature of the problem. It's not just candidates stretching the truth on a CV anymore – it's systematic fabrication and AI is making it accessible to anyone. The hiring teams that are most at risk are the ones who haven't updated their screening process to reflect that reality,’ says Neil Rose, CEO of Referoo.

These aren't edge cases. For mid-market and enterprise organisations running high volumes of hiring, the probability of encountering a fraudulent candidate in any given quarter is no longer negligible.

Why can't CVs and video interviews detect candidate fraud on their own?

The problem with relying on CVs and interviews as your primary screening tools is that both are self-reported and largely unverified. A candidate controls every word on their resume and, increasingly, can control what you see and hear in a video format too.

One-way video interviews, while efficient, are particularly vulnerable. There's no guarantee the person on screen is the applicant, no way to verify employment history in real time and no connection to an actual past employer who can speak to performance.

Even structured panel interviews, done well, are subject to skilled candidates who have prepared well or, in the worst cases, misrepresented who they are entirely.

How do reference checks protect against AI-enabled candidate fraud?

A properly conducted reference check is one of the few hiring steps that requires a real human relationship with a real employer to be meaningful. It can't easily be faked at scale and that's exactly what makes it valuable.

When a referee is contacted through an automated system that asks structured, role-specific questions, the barriers to fraud multiply quickly. Referoo also tracks IP addresses against referee responses, adding another layer of detection for candidates who attempt to complete their own reference checks.

‘A CV is a document the candidate writes about themselves. A video interview, even a live one, increasingly can't be taken at face value. But a reference check, when it's done properly, requires someone else to vouch for that person, and that's a fundamentally harder thing to fake. That's where the value sits,’ Rose says.

That doesn't mean reference checks are foolproof. But done properly, they add a verification layer that most other hiring steps simply don't.

What other checks should sit alongside reference checks to reduce fraud risk?

Reference checks are most effective when they sit within a broader screening process rather than operating in isolation. Two checks that address different dimensions of candidate fraud are work history verification and identity verification.

Work history verification goes beyond what a candidate has self-reported. Rather than taking employment dates and job titles at face value, it cross-references claimed history against verifiable records, surfacing gaps, inconsistencies or fabrications that a CV review and interview process are unlikely to catch on their own.

Identity verification addresses a more fundamental question: Is the person you're assessing actually who they say they are? With deepfake technology now capable of altering appearance and voice in real time, confirming identity through a verified document check adds a layer of certainty that a video interview alone can no longer provide.

Each of these checks is looking at a different aspect of a candidate's credibility. Used together, they make it significantly harder for a fraudulent candidate to pass through undetected at multiple points in the process.

‘No single check eliminates risk entirely, but layering reference checks with work history verification and ID checks creates a process that's genuinely difficult to game. Each one is looking at a different dimension of the candidate's credibility. Together, they give your team something defensible and that matters, whether you're protecting your business or just trying to make a good hire,’ Rose adds.

How can hiring teams reduce candidate fraud risk right now?

If you're reviewing your screening process with fraud risk in mind, a few practical steps are worth considering:

• Audit your current checks. Are you verifying employment history independently, or relying on what the candidate has told you?
• Move to structured, automated reference checks. These are harder to manipulate and generate more consistent data than ad hoc phone calls.
• Use a platform with built-in fraud detection. Purpose-built tools can flag suspicious patterns, duplicate referee details and other red flags your team might miss.
• Add ID verification before or alongside your reference check. Knowing who you're screening is a prerequisite for everything else.
• Treat inconsistencies as signals. If a candidate's CV, interview answers and reference check responses don't align, that's worth investigating.

Candidate fraud won't be solved by any single step in a hiring process. But the organisations that are most exposed are those treating screening as a formality rather than a genuine risk-reduction exercise.

See how Referoo Hub supports a more fraud-resistant hiring process

Referoo Hub brings reference checks, work history verification and ID checks into a single platform, making it easier to run consistent, defensible screening at scale.

Book a demo to see how Referoo Hub can support your team.