Recruitment Is Teaching Gen Z That Work Is a Transaction

When a young person's first encounter with an employer treats them as a data point, we shouldn't be surprised when they treat the job opportunity the same way.

For most young people, a graduate scheme or apprenticeship application is their first real exchange with the world of work. It's where they find out what employers value, how they'll be spoken to, and whether anyone is actually listening.

Right now, the answer they're getting is not a positive one. Somewhere between the online form, the automated screen and the one-way video interview, recruitment stopped being a conversation and became a transaction. Early talent are the ones being impacted most, given its their first experience learning what work means and how much they are valued.

Volume broke the relationship before AI arrived

It's tempting to blame the algorithm, but the pressure started with the gradual increase in candidate numbers. The Institute of Student Employers (ISE) latest student recruitment Survey puts competition at 141 applications per graduate vacancy, climbing to 172 in London. Two decades ago, the figure was 38.

The increase in the volume of applications over the years can be attributed to easier and faster application processes, increased access to higher education, more accessible criteria such as removing degree grade requirements and the increase in remote roles that makes the talent pool more global. However, there was a 59% increase in the number of job applications from 2022/23 to the following year. This was as we emerged from COVID and the 'great resignation' that followed, but it was also when generative AI took off.

No recruitment team can nurture a human relationship with that many people for each application. So the process was rebuilt around efficiency; forms that standardise, filters that sort, assessments that score. AI didn't invent that logic, but it did accelerate it.

The candidate didn't change. The design of recruitment did.

The person becomes whatever the system can read

Once a process is built for efficiency as its goal, the person applying is reduced to the parts a machine can parse. A keyword match. A percentile on a gamified test. A score attached to a recorded answer given to a webcam, with no one on the other side. An AI summary of all those data points wrapped up into a neatly packaged overview for decisions to be made. Decisions that impact human lives, where the individual has no oversight on the evidence being presented.

The Information Commissioner's Office audit of AI recruitment tools offers a glimpse of what sits behind that screen. The regulator made nearly 300 recommendations to the providers it reviewed, and found instances of tools inferring characteristics such as gender or ethnicity from a candidate's name.

At least one provider's accuracy benchmark amounted to a requirement to perform better than random chance. That's a pretty low bar to aim for when we are talking about deciding who gets an opportunity to be a future leader.

Candidates are rarely told any of this. Greenhouse's 2026 UK research flags candidates being evaluated by AI without being clearly told beforehand. Being assessed is fair. Being assessed without knowing how, by what, or against which criteria is something else entirely. If AI is being used, then it should be easier for employers to be transparent about how the process works, what criteria are used and how does it inform recruitment decisions.

Built for the buyer, not the user

Most AI in recruitment isn’t designed with candidates in mind, because candidates aren’t the customer. The employer pays for the tool, the employer sets the success measures, and the vendor builds to meet them; time saved, cost per hire, applications processed per recruiter. Those are the numbers on the sales deck, so those are the numbers the product is optimised for.

You can see it in how the technology is sold. Greenhouse promotes AI as cutting UK recruiters’ manual screening time from 23 hours to under three per placement. The ISE notes employers are increasingly drawn to AI as a way to manage application volumes and reduce costs. Neither goal is unreasonable. But notice who’s missing from both; the person on the other side of the screen.

In any other product, the user is the person who has to live with the experience, and good design starts with their needs. In recruitment, that person is the candidate, and they’re the one least consulted when it comes to incorporating AI functionality. Few tools are tested for whether a one-way video interview leaves a nervous 21-year-old feeling assessed or surveilled, whether a rejection message explains anything that helps the person develop and grow, or whether a gamified test makes sense to someone who’s never seen one before.

Greenhouse found 70% of hiring managers trust AI to make faster, better hiring decisions, while only 8% of jobseekers think it makes hiring fairer. That gap is what you get when one side of a process is designed for and the other is designed around.

Good products are built for, and centred around, all the key people who use them. Understanding their different needs and meeting them. Most recruitment AI is built around the people who pay for it.

Silence is the feedback early talent hear most

The transaction is clearest at the end of the recruitment process, when nothing comes back. Indeed found UK jobseekers hear nothing from more than half of the roles they apply for, and 86% said they'd rather receive a rejection than silence. A Tribepad survey of 2,000 UK jobseekers found close to two-thirds had been ghosted by a recruiter, with more than four in ten saying it took weeks or months to recover.

For an experienced professional, silence is irritating. For a 21-year-old sending their first applications, it's formative. Without feedback there's nothing to learn from, only a vague sense of having been assessed but not being successful. That can have a profound impact on a young person's wellbeing and sense of self worth.

The cost isn't evenly spread. Candidates who've had coaching on how automated assessments work, or can afford premium tools that game keyword filters, learn to perform for the machine. Those who haven't tend to read silence as a verdict on themselves rather than on the process. An opaque system doesn't just reject people quietly. It quietly rewards those who already know how it works due to their privilege.

Transactions run both ways

Employers tend to miss that Gen Z are paying close attention, and adapting. Greenhouse's multi-market research found 62% of entry-level jobseekers had lost trust in hiring. Trust is a fundamental cornerstone of building any strong, positive relationship making it a worrying statistic.

So, in return, Gen Z have changed how they play the game. The ISE found 92% of employers suspected or confirmed students using AI in assessment and selection, up from 85% the year before. Indeed found 55% of UK workers think ghosting an employer is fair play, given employers do it to them.

This is the transactional bargaining working exactly as designed, albeit unintentionally. If the employer reduces the candidate to a data point, the candidate optimises themselves into a better data point. If the employer goes silent, the candidate feels no obligation to stay in touch. Nobody is behaving badly. Everyone is behaving rationally inside a system that has removed every reason to behave otherwise.

Then the same employers wonder why their graduate intake doesn't feel loyal. The ongoing impact to the relationship between early talent and employers is lasting, based on how the person was treated and felt at the first point of contact.

Automate the admin, not the relationship

The fix isn't to strip AI out of recruitment. At 141 applications per role, some automation is inevitable and, used well, it can free recruiters up for the parts of the process that matter. The real question is what the technology is for. Buying back human time, or removing humans altogether?

That starts with telling candidates where AI is used, what it's assessing, and who makes the final call. The ICO already expects that transparency, so this is less a stretch goal than a baseline. It means closing every loop, because an automated "no" with one honest line of context is worth more than weeks of silence. It means putting a person at the point of decision, not just at the point of offer. And it means treating recruitment as the first stage of onboarding rather than a filtering exercise that happens before the relationship begins.

Because it already has begun. The application process is the first piece of company culture a young person experiences, long before induction day. Employers shouldn't just ask their new employees about their recruitment experience, but also ask all the unsuccessful candidates to really gauge how effective the process is.

The new terms are being set now

Every early careers team talks about attracting talent that's engaged, curious and committed. Then the first thing that early talent experience is a process that treats them as a row in a database and doesn't bother to acknowledge them unless they get to the next stage.

Gen Z aren't failing to understand the world of work. They're reading it accurately.

You can't recruit people as data alone and expect them to show up as humans who feel connected, committed and loyal to the employer.