Tom Cornell at HireVue explores some of the challenges posed by candidates cheating in during the hiring process and offers some solutions
In the competitive world of hiring, cheating has become a significant concern for employers. Extensive research has been conducted by leading AI hiring companies to understand the prevalence and impact of various cheating behaviours in the hiring process. This article summarises these findings and offers strategies to mitigate these challenges.
Types of cheating behaviours
Artificially enhanced answers: with the rise of AI tools like ChatGPT, there’s growing concern about candidates using these technologies to artificially enhance responses. However, new research suggests that the risk of hiring unqualified candidates due to ChatGPT usage is relatively low.
Shared responses: another form of cheating involves candidates sharing answers to interviews or assessments. Across a series of studies in 2023 and early 2024, HireVue analysed 10,000 transcribed candidate asynchronous interview questions. The results were encouraging, with less than 1% of responses flagged for high similarity.
Impersonation: perhaps the most concerning form of cheating is impersonation, where someone other than the candidate completes the interview or assessment. This issue can be particularly prevalent in certain geographical areas.
Detecting cheating behaviours
Recognising artificially enhanced answers often involves identifying irregular speech patterns, more formal language, and a lack of filler words typically present in natural speech.
For shared responses, AI platforms can employ sophisticated algorithms to detect similarities between candidate answers that are robust to attempts to hide these behaviours.
Strategies to reduce cheating
Multi-stage workflow validation: implementing a multi-stage hiring process that includes a live interview can help verify a candidate’s abilities and authenticity.
Code challenge tools: for technical roles, AI code challenges include similarity scores and playback features to detect potential cheating.
Follow-up questions: asking candidates to explain their thought process or revisiting answers given in previous steps can reveal whether they truly understand the concepts or have used artificial assistance.
Time-limited questions: while spontaneous questions can limit the use of external resources, they may also increase candidate anxiety and reduce overall effectiveness, making balance important.
Smart question selection: crafting questions that require candidates to apply their skills and experiences, rather than recite information, can neutralise the impact of AI tools.
Frequent question updates: regularly changing interview questions can prevent candidates from sharing effective responses.
The prevalence of cheating
Despite concerns, research indicates that cheating is not as widespread as some might fear. In a new study, only a small percentage of candidates showed evidence of using shared responses or other forms of cheating. Even when cheating was detected, these candidates often performed poorly and were not selected to move forward in the hiring process.
The future of hiring integrity
As technology evolves, so too will the methods used to cheat in hiring processes. While cheating in hiring remains a concern, the data suggests that it’s not as pervasive or effective as many fear.
By implementing smart hiring strategies and leveraging advanced tools, employers can maintain the integrity of their hiring processes while ensuring a positive experience for honest candidates.
The key to successful hiring lies in a multi-faceted approach that combines technology with human insight. By asking the right questions, utilising multi-stage processes, and staying vigilant, companies can effectively identify top talent while minimising the impact of cheating behaviours.
As the hiring landscape continues to evolve, it’s crucial for organisations to stay informed about the latest trends and tools in candidate assessment.
Tom Cornell is Senior IO Psychology Consultant at HireVue
Main image courtesy of iStockPhoto.com and Orbon Alija
© 2024, Lyonsdown Limited. Business Reporter® is a registered trademark of Lyonsdown Ltd. VAT registration number: 830519543