How Deepfake Technology Challenges Video Interviews

Video interviews are now essential in recruitment, providing convenience, efficiency, and the ability to evaluate candidates anywhere in the world. Yet, organisations face a significant challenge: candidates increasingly use AI-driven 'deepfake' technology to manipulate their interview performance undetectably.

Deepfake refers to sophisticated AI technology capable of creating realistic and convincing fake videos or altering live videos in real-time. In recruitment contexts, candidates leverage deepfake tools to simulate genuine eye contact and spontaneity, even while covertly reading pre-generated answers from AI tools like ChatGPT.

Initially, interview cheating involved candidates visibly reading scripted responses.

Recent surveys suggest an alarming 25% of candidates may already employ deepfake tactics in their video interviews.

The proliferation of deepfake technology brings critical issues:

• Compromised Authenticity: – Candidates convincingly fake interview responses, undermining the reliability and fairness of the recruitment process.
• Increased Hiring Risks: – Employers risk hiring candidates who misrepresent their skills, abilities, or genuine responses during interviews.
• Diminished Interview Effectiveness: – The inherent value of video interviews—observing authentic reactions and interactions—is significantly undermined, making it challenging to assess a candidate's true suitability.

• Effective Deterrent: – Simply informing candidates that verification tests will be conducted substantially reduces the incentive to cheat.
• Reliable Detection: – Verification tests accurately identify cheating by measuring performance gaps between remote and supervised assessments.
• Future-Proof: – As AI models grow even more powerful, verification tests will remain a viable solution, consistently identifying discrepancies no matter how advanced AI becomes.

Addressing the deepfake challenge requires strategic adjustments in recruitment processes:

• Utilise Robust Early Screening Methods: – Prioritise structured assessments like psychometric tests, situational judgement tests, or skills-based evaluations early in the recruitment process. These assessments are far less susceptible to manipulation.
• Late-stage Video Interviews: – Reserve video interviews for later stages, where fewer candidates remain, allowing for more comprehensive verification procedures, including in-person or supervised virtual follow-ups.
• Candidate Transparency and Education: – Clearly communicate to candidates about the usage of advanced detection methods and consequences associated with deepfake cheating. Transparency can significantly deter potential misuse.

The evolving landscape of AI technologies, including deepfake capabilities, necessitates continuous vigilance and adaptability. Organisations must:

• Stay informed about technological advancements in AI and deepfake detection.
• Invest in AI-driven anti-cheating software and real-time analysis tools to help identify irregularities during interviews.
• Regularly review and update recruitment procedures to ensure effectiveness against emerging threats.

By proactively addressing the deepfake threat, organisations can maintain integrity, authenticity, and trustworthiness in their recruitment processes, ensuring they select genuinely qualified and suitable candidates.