AI in recruitment: where the gains stall determines if it works

A screening that used to take an hour now takes fifteen minutes with AI. A job description you once typed out yourself can now be generated in a few minutes. The time saved is real and measurable. The question that remains underexposed in most recruitment articles is: where does that time go?

That is exactly what it’s all about. Not whether AI helps—that’s already a given—but who actually benefits from it.

Knowing more about AI doesn't necessarily make you feel more at ease

Research from Statistics Netherlands (CBS) reveals something striking: workers who already use AI are more likely to think their jobs could be (partially) taken over than those who don't. 56 percent versus 37 percent. You would expect the opposite; becoming familiar with a technology should temper concerns about it. Apparently, it works the other way around.

For recruiters who use AI for screening or scheduling, this is relatable. You see exactly what it saves you, and precisely because of that, you see most clearly what is changing in your profession. Of all Dutch adults, 75 percent think AI will make jobs disappear, 64 percent expect a loss of knowledge and skills among staff, and 57 percent simultaneously predict higher productivity. These thoughts are not contradictory. They are proof that people see both what they gain and what they lose.

The gains don't automatically go to those who use AI

According to Indeed, the impact of AI varies globally, depending on local labor needs, regulations, and culture. But the core of their analysis lies in one choice that recurs everywhere: do companies reinvest the time saved into their people, or does it primarily go toward profit? Kyle M.K., Talent Strategy Advisor at Indeed, puts it this way: we will need to rethink how we value people, not just for their productivity, but also for the quality of their ideas and the way they interact with others.

Translated into recruitment, this becomes very tangible. Fifteen minutes saved on screening can mean two things. You fill more vacancies in the same amount of time, and the gain goes to revenue. Or, you spend that time on the conversation with the candidate—the questions you normally didn't have room for. Both are legitimate choices. The problem is that the first option often becomes the default, simply because no one consciously chooses otherwise. The time disappears into higher volume, not into more attention.

That is also the difference between AI that makes recruitment faster and AI that makes recruitment better. Both sound the same in a sales pitch. They are not.

What AI in recruitment won't take over, no matter how much time you save

No matter how much time AI saves, a few things in recruitment will always remain human work. Indeed specifically mentions them: assessing a candidate's potential, rejecting someone with care, and guiding a new employee through their first few weeks. No tool that intervenes in that process will change that, no matter how good the tool is.

Trey Causey, Head of AI Ethics at Indeed, is sharp on this point: he warns recruiters to be highly skeptical of AI recruitment software providers who claim they have solved complex recruitment problems with AI. That is a useful warning, because precisely because AI saves so much time in screening and scheduling, the temptation is great to think that the rest of recruitment can also be taken over by AI. Assessing a candidate, rejecting them with care, onboarding: that remains human work, no matter what a vendor promises.

There is another side worth mentioning: candidates are using AI just as much, for example, to write resumes and cover letters. Detection tools for this work poorly; OpenAI's own tool misses 74 percent of AI-generated text. What does help is simply reading carefully: a cover letter that sounds remarkably polished and monotonous throughout often reveals more about AI usage than any detection tool ever could.

The question you really should be asking yourself

The question "will AI replace recruitment" leads nowhere, because the answer has long been clear: no, not the core of the profession. The question that actually matters is rarely asked: where do you let the time you save land? In more volume, or in the part of recruitment that will always remain human work?

That answer determines not only how much time you save. It determines whether AI makes your profession better, or just faster.

At Collab, we deal with that question every day: how to deploy AI in a way that benefits recruitment without losing the human element. Are you running into this yourself? We’d love to think along with you.

‍