Over the next year, early talent teams are gearing up for what can only be described as an application avalanche. With generative AI tools now embedded into every corner of the job‑seeking experience – from CV creation to automated interview helpers – it’s never been easier for candidates to apply quickly, widely… and sometimes, not entirely honestly.
While these technologies genuinely help many candidates put their best foot forward, they also open the door to sophisticated forms of candidate fraud that HR and Talent Acquisition teams weren’t dealing with even 18 months ago. The good news? With a mix of awareness, process discipline, and the right tools, teams can stay confidently ahead of the trend.
Below are the some of the most common types of candidate fraud surfacing today, paired with practical signals and mitigation strategies to help hiring teams navigate this evolving landscape.
- Real‑Time Facial Deepfakes (AI Masks)
- What is it? AI-generated images or videos that convincingly imitate real people’s faces or actions.
- What does it look like? A candidate’s face appears slightly “off” – flat lighting, mismatched lip movements, or an oddly static background.
- What should you do? Ask the candidate to adjust their camera, perform a quick live action (like waving), or share their screen for a brief task.
- Synthetic Voice Generation
- What is it? AI‑based voice replication can now recreate someone’s vocal traits with striking precision, modelling elements like rhythm, accent, tone, and even subtle breaths to produce audio that closely mirrors the original speaker.
- What does it look like? A voice that feels overly polished, synthetic, or oddly emotionless.
- What should you do? Throw in spontaneous, conversational questions. Ask for examples about experiences not easily scripted. Record segments where the tone shifts or where responses sound processed.
- Off-Screen Browser Support
- What is it? Candidates secretly using external resources, like hidden browser windows, a second monitor, or off‑camera devices, to look up answers during an interview or assessment.
- What does it look like? Eyes darting to another screen, typing sounds while they speak, or suspiciously accurate technical answers.
- What should you do? Ask for a quick screen share or move to a live, practical exercise.
- Lip‑Syncing
- What is it? Candidates silently watch a pre-recorded response while mouthing along, making it seem like they’re answering in real-time.
- What does it look like? A candidate moves their mouth minimally – or not at all – while someone/something else supplies the answers off‑camera.
- What should you do? Request narration of an on-screen task or ask the candidate to speak while performing a simple movement. Record & make note of timestamps where speech doesn’t match mouth movements.
- AI Interview Copilots
- What is it? AI Interview Copilots are assistants that help candidates by providing real-time guidance, suggested responses, and structured feedback to improve performance. These tools can analyse questions, generate optimised responses and offer coaching as well.
- What does it look like? Perfectly structured, overly general responses often with a noticeable delay.
- What should you do? Use “prompting questions” that build on previous answers. Ask something unexpected or scenario-specific questions which would be difficult for AI to answer.
- Human Imposters
- What is it? Old-school fraud is alive and well – someone else completes part of the process (i.e. video interview, phone interview, virtual interview).
- What does it look like? The candidate’s experience, appearance or behaviour doesn’t match their documents, online presence or earlier interactions.
- What should you do? Request ID verification or speak with previous managers or references. Look out for any mismatched shared work experience and/or CV details.
- Sophisticated Multi‑Layered Attempts
- What is it? Some candidates combine several methods for a more convincing façade.
- What does it look like? A mix of red flags – voice anomalies, perfect answers, and off-camera sounds.
- What should you do? Use multi-step verification and multiple interviewers and collect evidence across platforms.
How to Handle Suspicious Behaviour Carefully and Professionally
Recruitment teams should balance vigilance with fairness. The goal isn’t to confront candidates in the moment – it’s to document, assess, and escalate appropriately.
When something doesn’t feel quite right in an interview, recruiters should keep the conversation neutral and avoid addressing concerns in the moment. Ending politely and taking time after the call to review any odd moments – whether timing delays, inconsistencies, or technical anomalies – helps maintain fairness and objectivity. Noting approximate timestamps and gathering supporting materials such as recordings, platform alerts, and profile information ensures you have a complete picture before drawing conclusions.
From there, concerns should be documented factually in the ATS using careful, non‑accusatory language and escalated through the appropriate internal channels, such as HR, Security, or Legal.
Many organisations are also reverting to in-person processes, partly to elevate the candidate experience, but also to aid in verifying identity. This balanced approach protects candidate experience while maintaining the integrity of the hiring process.
The Growing Toolbox for Fraud Prevention
The best defence blends technology, process, and people.
- Assessment tools with anti-cheating mechanisms (i.e. disabled copy & paste, active monitoring of click rates, eye movements and recording functionality etc.)
- Interview platforms with built-in fraud detection
- Panel interviews to reduce the chance of missed cues
- Supervised retests for technical or problem-solving skills
- Identify verification including government ID checks and employment verification
- Interviewer training to recognise emerging fraud patterns
- Dedicated fraud prevention groups who are responsible for sharing insights and best practice tips across the organisation
- In-person touchpoints to minimise online manipulation
The Takeaway – Vigilance Without Paranoia
Candidate fraud is evolving quickly; but so are our tools, processes, and collective awareness. Most candidates act with integrity, and TA’s role isn’t to treat everyone with suspicion. Instead, it’s about being informed, observant, and ready to act when things don’t quite add up.
The organisations that successfully navigate this new era won’t just rely on technology; they’ll empower their teams, modernise their processes, and maintain a culture of fairness and trust.
