The impact of generative AI on hiring is now impossible to ignore. Here are some eye-opening recent examples of the impact, and the very real struggle employers are now grappling with:
One-way video interviews are becoming a sea of script-reading, as candidates recite ChatGPT-generated answers. For example, one major professional services firm saw its one-way video interview pass rate plunge from 50% to 13% year-on-year because so many applicants submitted identical ChatGPT responses - with significant resource implications on their hiring process.
Many employers are also seeing application numbers surge by over 100% year-on-year, which in itself creates challenges - but it's more complicated than that.
The vast majority of this increase coming from male applicants. This has led in some cases to shifts from a previous 50/50 gender split in applications, to a split as high as 75/25 (in favour of males).
Candidate survey data explains this trend, revealing that men are significantly more likely to use AI to make mass job applications - and to ignore instructions not to use Gen AI. Female-to-male offer ratios are now taking a hit, undermining gender equality.
Meanwhile, near-perfect application form responses are flooding in, rendering traditional application questions almost worthless for screening. GPT-o1 and o3 can produce sophisticated replies on any topic - technical, motivational, or otherwise - making it difficult to assess genuine job fit.
Yes, I fully support the idea that smart use of AI can boost productivity. But when generative AI dominates the hiring process, it masks genuine talent and potential. Assessing behavioural strengths such as resilience, empathy, and learning agility becomes increasingly difficult if using application forms, video interviews, or conventional psychometric tests.
Back in 2023, I stood on stage at the Institute of Student Employers conference and talked about generative AI’s likely impact on hiring. I also shared survey data demonstrating how EDI outcomes were at risk. Some assessment providers called it “scaremongering,” but it was a straightforward logical deduction - AI is rapidly becoming both easier to use and more powerful, so the impact we now see in hiring processes was entirely predictable. That reality is no longer in question.
The consequences go beyond the integrity of assessments. There are now quantifiable, negative EDI outcomes driven by demographic differences in AI usage. This was entirely foreseeable, but much of the assessment industry chose to ignore the warning signs, accused others of scaremongering, and left employers to pick up the pieces.
The simple truth is that conventional hiring processes - including one-way video interviews and most standard psychometric tests - are not designed for the age of advanced generative AI.
That has been true since I first highlighted it two years ago, and it remains just as true today. The time for meaningful change is now.