What is candidate fraud and how employers can prevent it

All I needed to do was pick one of the many free platforms available, upload a photo of myself along with a short audio clip of me talking and hey presto. I was able to create a video of someone who looked exactly like me and who could enthusiastically recite any given text, in any language, using my voice and intonation.

Admittedly, someone who knows me well would probably get suspicious interacting via video with New AI Me. But if New AI Me was being interviewed online for a job the employer probably wouldn’t have a clue that all may not be as it seems. Especially if that interview was an asynchronous one that was being assessed by an algorithm looking for keywords. AI candidate meets AI interviewer and it’s love at first sight

This AI avatar could have already submitted a fake CV with fabricated work experience and imaginary qualifications. As the employer you are loving this CV and the candidate’s responses to your questions….but who are you actually interviewing? And when you offer them the job who will show up on Day 1?

Remote interviewing makes it easier for candidates to manipulate the process and while using AI-generated deepfake technology for candidate fraud is still rare, it does happen.

The most commonplace types of candidate fraud:

1. Fake CV and/or LinkedIn profile: employment, education or skills are fabricated or wildly exaggerated. Does the person even exist?
2. Real-time assistance during the interview: using AI platforms to receive answers or support during the interview.
3. Deepfake Candidate: using software to alter voice or appearance in video interviews.
4. Professional Interview Candidate: paying someone who is an interview professional to attend and ace the interview in place of the “real” candidate

Mitigating candidate fraud

It’s real. Here are 7 actions you can take to spot and reduce the occurrence of fraud in your candidate selection and assessment process

1. Train your interviewers to notice red flags – eg excessive pauses, highly polished answers or inconsistencies between what they said in their CV/application and what they are telling you in the interview
2. Consider having a policy of recording interviews, either solely audio or audio and video. It helps with reviewing and analysing candidate responses post-interview and is also good for identifying inconsistencies if required. Tell the candidate in advance, explain the purpose of the recording and that it will be stored in line with your GDPR policies. Also be clear with candidates that your company has a policy that prohibits candidates from recording interviews
3. Avoid asking open-ended questions – they are very easy for the AI to answer. Ask for specific examples of real experiences and then ask follow-up questions….. and keep asking them. Ensure that your interviewers know how to probe an initial response
4. Insist on cameras on. When inviting the candidate to interview tell the candidate that this is a non-negotiable policy. “Sorry my camera just broke today” or “My wifi is really bad so I better turn off my camera” could be 100% legitimate or could be an attempt at candidate fraud.
5. Ask for proof of ID: when inviting the candidate to interview tell the them that they will need to show proof of their ID on camera before the interview starts. Don’t feel awkward about doing it. Legitimate candidates will be happy to provide it. PS a LinkedIn profile is not proof of ID!
6. Don’t just tell, show: Include work-based tests or practical tasks, preferably in real-time, as part of the candidate evaluation process
7. Re-introduce in-person interviews: while more time-consuming they are the best way to reduce candidate fraud.

Download our Candidate Fraud Prevention Checklist

Keep your recruitment processes safe from attempts to deceive your organisation with fictitious profiles or AI generated deepfake candidates.