Good conversational AI interview screening, according to candidates
Jump Links placeholder, DO NOT REMOVE
- Nine in 10 U.S. candidates are open to an AI interview screen, and only 7% would refuse one outright – resistance isn’t the problem, poor design is.
- Trust comes down to five specific design choices: whether the screen listens, follows up on vague answers, answers candidate questions, allows time to think and is upfront about AI and human review.
- A poorly designed AI screen can shape a company’s reputation before a hiring decision is even made, since it’s often a candidate’s first substantive interaction with an employer.
- Greenhouse Voice AI is built around dynamic follow-ups, full transcripts and human review on every decision – never scoring tone, accent or confidence.
A candidate clicks into an interview link and finds no interviewer on the other end, just a voice asking the next question. Maybe they’re relieved it isn’t another resume form to fill out. Maybe they’re bracing for a script that won’t adjust no matter what they say.
Either reaction is fair. Hiring is already stuck in what Greenhouse calls an AI doom loop. “For the first time I can remember, neither side thinks hiring is working,” said Sharawn Tipton, Chief People Officer at Greenhouse. “Candidates send applications into a black hole, while employers sift through AI-polished resumes that all look the same. The application used to be a front door. Now, too often, it’s a dead end.”
Candidates need a system that actually works and gives them a fair shot. Bolting bad AI on top only hurts them and employers.
– Sharawn Tipton, Chief People Officer at Greenhouse
Sharawn’s distinction, bad AI versus AI built with intention, comes down to design. Here’s what candidates actually feel about AI interview screens, what sets those experiences apart in earning or losing trust and what’s at stake for your employer brand.
Most candidates are open to AI screening when it’s designed well
Recruiters sometimes assume candidates grit their teeth through AI interview screens, waiting for a human to show up instead. Greenhouse’s 2026 Future of Hiring Candidate Report, which surveyed more than 2,800 job seekers across the U.S., U.K., Ireland, Germany and Australia, suggests something else: most U.S. candidates are reacting to how poorly some employers have implemented AI candidate screening, not to the format itself. That tracks with how the doom loop usually plays out in practice: a lot of these screens were built like scripts with a microphone attached, not like interviews.
Nine in 10 U.S. candidates say they’re open to an AI interview, and only 7% would refuse one outright. They’re also 2.7x more likely to take an AI interview than to withdraw from a process where no human interviewer is available. Nearly half (48%) have already taken one in the past year, and 70% of those say it let them demonstrate their skills well.
What tips a screen from bad to good is largely about what candidates are told going in. Nearly half (49%) of U.S. candidates want to know a human will review their responses, 40% want to know how they’ll be evaluated and 88% want at least one legal requirement that makes employers upfront and accountable about AI in interviews, with advance notice the leading ask at 55%. Get those basics right, and there’s a real opening most employers aren’t using yet.
Candidates see a new opportunity to be heard
Consider what candidates are dealing with globally: 81% across the markets Greenhouse surveyed were rejected from a job in the past year without speaking to anyone at the company, human or AI. Among U.S. candidates specifically, 96% wish they could show employers something their application simply doesn’t capture – more than half (52%) point specifically to their ability to learn and adapt.
“AI can polish a resume or fire off hundreds of applications, but it can’t show an employer how someone thinks or learns,” said Ophir Samson, Head of Voice AI at Greenhouse. “A structured conversation early in the process could give the right candidates that chance. But it only works if candidates understand how they’re being evaluated, and know a real person will review their responses. Trust must be built into the process from the start.”
That’s the opportunity a good AI interview screen represents: a chance to actually hear a candidate instead of filtering them out on paper, and to free up recruiter time for the parts of the job that actually move hiring forward – the hallmark of a strategic recruiter rather than a screening machine.
What good AI interview screening looks like in practice
Trust comes down to specific design choices candidates can spot immediately. Here are five traits that build it, and a few patterns that quietly erode it.
It listens, instead of queuing up the next question
A good AI candidate screening tool tracks what a candidate actually said and responds to it: mention leading a migration project, and it asks about team size or what broke along the way. A careless one moves to the next scripted question regardless, even one on a topic the candidate already covered unprompted. It also makes for a better transcript afterward – a recruiter reviewing it later sees not just what a candidate said, but how they handled being pressed on it.
It follows up on vague answers instead of accepting them at face value
A thin answer – a one-line response to a behavioral question, or one that dodges the specifics – gets a follow-up from a good screen: what role did they play, what was the outcome. A good follow-up can be as simple as asking for one specific number, one decision they made or one thing that didn’t go to plan. A weaker screen accepts whatever comes through the microphone and moves on. The follow-up does double duty: it’s also part of how AI candidate screening catches fraud, since a scripted or AI-generated answer tends to fall apart under a specific question in a way a genuine answer doesn’t.
It lets candidates ask questions and get straight answers
Good AI interview screening tools run in both directions. A candidate should be able to ask about compensation, team structure or what the role actually looks like week to week, and get an answer grounded in what the hiring team shared when they set up the screen. A poorly designed one deflects, loops back to a canned statement or skips the moment entirely – telling a candidate plenty about the rest of the process before a recruiter ever follows up.
It doesn’t interrupt when someone pauses to think
Thinking out loud takes a beat, and a good screen can tell the difference between a candidate who’s finished talking and one who’s still forming a thought. A screen that talks over a pause trains candidates to rush and under-answer – usually just a matter of tuning the screen to tolerate a few extra seconds of silence, not a technology limit. That penalizes non-native speakers and anyone who thinks before committing to an answer, the exact group a considered question is meant to surface.
It’s upfront about what’s AI and what’s recorded – and is clear that humans own the decisions
Candidates should know three things before they start: they’re talking with AI rather than a person, a person will review their responses before anyone decides anything and what happens to the recording afterward. Ideally, you’d share transparent guidelines on your company’s AI use in hiring, but if you don’t have that, a sentence before the first question and a line in the interview confirmation email usually covers it. Saying it plainly costs an employer nothing, but makes all the difference.
Get these five right, and AI interview screening stops being something candidates tolerate and starts being something they’re genuinely open to.
What’s at stake for employer brand
Get AI interview screening right, and candidates lean in. Get it wrong, and the fallout doesn’t stay contained to the thirty minutes a candidate spends on the call. Hiring is already a reputational moment, and a full pipeline doesn’t change that.
Even today, the candidates most worth hiring still have options, and the best recruiters know that they’re forming an opinion of the company as they go. AI screening raises the stakes because it’s often a candidate’s first substantive interaction with an employer – the first time anyone, or anything, representing the company actually responds to them. That puts candidate experience on the line before brand reputation even enters the picture.
A bad screening experience can turn that first interaction into one more bad story: on review sites, in group chats with other job seekers, in conversations with people who might have been the company’s next strong hire.
No employer gets to approve that story before it spreads. The only real lever left is making sure it was worth telling in the first place.
The bottom line
Candidates aren’t asking employers to abandon AI interview screening. The stats say the opposite: most are ready for it, and plenty would rather talk to a well-designed screen than watch an application disappear into an unread queue. What they’re asking for is a screen that behaves like the hiring team is still present: listening, following up, answering honestly and telling candidates upfront what’s happening and why.
This is the standard Greenhouse built into Voice AI: dynamic follow-ups grounded in what a candidate actually said, full transcripts recruiters review before any decision gets made and no scoring of tone, accent or confidence. So candidates can finally feel seen – and get hired.
Curious what a well-designed Voice AI screen sounds like for your own hiring process? Get in touch.
FAQs
What makes an AI interview screen feel well-designed to candidates?
Candidates notice when a screen listens and responds to what they actually said, follows up on vague answers instead of accepting them, lets them ask questions about the role and get a straight answer, and gives them time to think without interrupting. They also notice whether the employer told them upfront that they’re talking with AI and that a person will review their responses.
Are candidates comfortable with AI candidate screening?
Most are. Nine in 10 U.S. candidates say they’re open to an AI interview, and only 7% would refuse one outright. They’re also more likely to take an AI interview than to withdraw from a process altogether when no human interviewer is available.
What should employers tell candidates before an AI screening interview?
Candidates should know they’re speaking with AI rather than a person, that a human will review their responses before any hiring decision and what happens to the recording afterward. Transparency on these points is one of the clearest ways employers build trust in AI screening.
How does AI interview screening affect employer brand?
A poorly designed screen becomes part of how candidates describe a company afterward, the same way a ghosted application does. Since AI screening is often a candidate’s first substantive interaction with an employer, getting it right or wrong shapes reputation before a hiring decision is even made.
Should a human review AI interview screening responses?
Yes. Candidates consistently say they want a person reviewing their responses before any decision is made, and the hiring decision itself should stay with people, not AI. AI can surface signal from a structured conversation; a person still has to act on it.
