What Candidates Actually Object To in AI Screening

Ask a room of recruiters whether candidates dislike being screened by AI and most will say yes. It has become one of those things everyone knows. And because everyone knows it, it gets used as a reason to keep a human on the first call, whatever that costs.

I think the belief is mostly true of something else. The useful question is which kind of AI people had in mind when they formed it.

As far as I can tell, much of the case against AI in hiring was built on studies from 2022 and 2023 about asynchronous video interviews and automated resume parsing. The candidates in those studies had good reason to complain. Recording yourself answering questions for a blank screen is an odd and lonely experience. You get no reaction and no follow-up, and nothing tells you anyone is on the other end. If that's what AI screening meant, I'd dislike it too.

A live voice conversation is a different thing. The candidate talks, and what they say gets a response. If they mention something relevant, the next question picks it up. From their side it's much closer to a phone screen than to a video monologue or a form. People don't usually mind phone screens, as long as they happen soon. Nobody enjoys being assessed, but being assessed quickly is evidence that someone took your application seriously.

So I suspect what candidates really object to is older than AI. It's waiting. Think about it from the applicant's side. You apply on a Tuesday, and a recruiter calls back on Friday, or sometime the following week. Even if the call is pleasant, the delay has already told you where you stand, which is somewhere in a pile nobody had reached yet. Now imagine getting a call the same evening you applied. It's hard not to read that as a company that pays attention to the people who apply.

When you ask candidates directly what bothers them, the complaints are mostly about inconsistency: a 20-minute conversation followed by two weeks of silence, a friendly intake call followed by a generic rejection with no feedback, being told "we'll be in touch" and then never hearing back. None of that is about who ran the first screen. It's about whether the process seems organized and whether it keeps its promises. An AI that calls someone back within hours, late in the evening if that's when they applied, and then follows up when it said it would is giving a better experience than many candidates get from human recruiters.

The mistake is taking what we learned about async video and treating it as a verdict on all AI in screening. It's an expensive mistake, because it leads teams to turn down a fix for a problem they've misdiagnosed. What candidates were rejecting was the feeling of talking into a void. A four-day wait for a callback gives them the same feeling.

Where the wait comes from

Nobody designs a gap of several days between application and first contact. It's what you get whenever applications arrive faster than people can answer them. Applications come in at night and on weekends, while recruiters work business hours and carry other reqs. Suppose a high-volume campaign brings in 500 applicants in three days. Even a good team will reach some of them quickly and the rest late, and the ones reached late are the ones who come away with the worst impression.

Closing that gap is the reason we built Asendia. When someone applies, it calls them and runs a structured screen based on that role's criteria, following up on what the candidate actually says. What the recruiter finds in the ATS afterward is a written summary for each person, with excerpts of what they said, and because everyone was asked the same core questions, one summary can be laid next to another and compared. The applications that would have piled up over the weekend become a ranked shortlist by Monday morning. The part I care about most is that the 400th applicant gets the same prompt conversation as the first. Nobody falls through because a recruiter ran out of hours on a Friday afternoon.

How fast that first contact happens also shows up much later, at the offer stage. Candidates who felt answered stay engaged, and engaged candidates are more likely to say yes. I won't repeat that argument here, since the post on why offer rejection rates are climbing goes through it.

So when someone tells you candidates hate AI screening, ask which experience they mean. Usually, I think, what they're describing is silence. A human team with more applications than hours has that problem just as much as any software does.

Ready to transform your hiring strategy? Schedule a Demo with our founders today!