I recently took part in a workshop organised by the EY Foundation, exploring the impact of AI on youth employment. What struck me particularly was that, far from being enthusiastic about these technologies, the highly articulate people in the room were deeply suspicious and cynical about it. They saw AI-supported recruitment as a barrier to employment – one that devalued individuality and prevented them from demonstrating their talent.

The fundamental error in most AI-supported recruitment is that it is focussed on cost reduction and “efficiency” – at the expense of the user experience. The psychological contract between recruiter and potential employee has been broken. The applicant is no longer a human being – just a set of numbers. The algorithms do not support and enable people to show themselves and their unique talents. Instead, they reduce the experience to a lottery based on yes-no decisions that are opaque to applicants (and arguably to the human recruiters as well).

In a classic clip from the Office, David Brent demonstrates his technique for working through an enormous pile of job applicants. He arbitrarily splits the pile in two, then drops one into the bin. His rationale? “They are the unlucky ones and we don’t need unlucky people!” It’s uncomfortably close to using AI to whittle out on-line applicants.

There does not seem to be any evidence that existing AI supported recruitment leads to better candidate selection. Rather, the scales are tipped towards those that are best at gaming the system, using AI to create applications that apply the right buzzwords and fit the formula. How that is supposed to correlate with, for example, leadership competence is a mystery! The danger is that candidates, who pass initial screening are homogenous clones. Outlier candidates, who bring innovative, unconventional ideas and qualities – and are essential to a healthy organisation – are squeezed out. Moreover, because the AI focuses on roles rather than potential, there is an increased risk of packing the lower levels of organisations with people, who don’t have what it takes to become effective leaders at higher levels.

The disillusion with AI-supported recruitment is not just confined to the young. Experienced employees know that the way into new jobs is via networking. The human connection is what counts.

Then, there is the wider psychological contract. When we have a bad experience with a company, it impacts our emotional response to it. So, feeling let down by a company’s recruitment processes affects how we view them from the perspective of a potential or actual customer. If they don’t appear to care in one context, why should we expect them to do so in another? The basic trust equation has been broken. Or to put it another way, any job applicant, who feels they have not been treated with respect as a human being, is less likely to have much respect for the company concerned.

Equally troubling is the lack of evidence concerning the impact of these approaches on equality and diversity. Do they disadvantage certain neurotypes, for example? Is the process more favourable towards people, who are good at filling in forms, or people who prefer to have a conversation?

What can recruiters do to redress the damage? Some potential remedies include:

  • Making it very clear in job adverts how applications will be assessed.
  • Offering video-based advice on how to apply.
  • Using an AI persona of someone, who has been in the vacant role, to have a conversation with applicants, to give them an idea of what’s needed, what the highlights and lowlights of being in the role are.
  • Giving applicants useful feedback. The biggest grudge is hearing nothing or receiving only a curt rejection note.
  • Doing the homework – researching amongst the target audience how they prefer to apply; finding out from applicants how they experienced the process. (Contrasting the experiences of those, who pass each stage, with those, who don’t, can provide a lot of valuable information!)
  • Making the process more like an interview, where the AI can get to know the whole person.
  • Having a reputation as a good employer isn’t compatible with one as a lousy recruiter!
  • Viewing every unsuccessful applicant as a valuable resource – as a future employee, collaborator, customer or champion referring others – and treating them accordingly.

The bottom line is that the systems currently in use are mostly creating barriers in the relationship between organisations and job applicants. Used intelligently, AI could enhance that relationship. But it needs a complete rethink, because current approaches benefit neither the employer nor job applicants.

©️David Clutterbuck, 2026