What Candidates Actually Say After Talking to an AI Interviewer

There is a story the hiring industry keeps telling itself: candidates hate AI interviews. They feel cold. Robotic. Dehumanizing. Every LinkedIn comment section confirms it. Every conference panel nods along. We have written about why that framing is wrong before. Now we have the data to prove it.

Over a 30-day window, we collected continuous post-interview feedback from thousands of candidates across hundreds of organizations spanning staffing, skilled trades, logistics, healthcare, distribution, automotive, and tech. This was not a curated survey with a controlled sample. It was every single opt-in response collected immediately after a Joy AI phone screen, including the ones where the candidate was calling from a bus, had a dead microphone, or told the AI to make them a job offer (yes, that happened).

The result: 83% rated the experience Excellent or Good. More than 62% rated it Excellent outright.

That is not a best-case number. That is what happens when you leave the feedback form open for everyone and count every response, including the edge cases and the sarcastic ones.

What hiring leaders get wrong about AI interview candidate experience

The conventional wisdom about AI interviewing is built on vibes, not data. Hiring leaders hear "AI phone screen" and project their own discomfort onto candidates. They picture a stilted phone tree. They imagine frustrated applicants hanging up. They assume the candidate experience will be a liability.

The data says something different.

50% of respondents left a written comment. That is an unusually high engagement rate for opt-in post-interview feedback. And when you read through those comments, one theme dominates everything else: candidates kept saying Joy felt like talking to a real person.

Not a chatbot. Not an automated system. A person.

One candidate said it felt like talking to a human manager directly. Another described it as attentive and quick with responses. A third said it did not feel mechanical at all. These comments were completely unprompted. Nobody asked candidates to compare the experience to a human conversation. They just did.

That matters because the biggest barrier to AI screening adoption has always been emotional, not functional. The technology works. The question hiring leaders keep asking is whether candidates will accept it. This data answers that question clearly. And it aligns with what we found in our broader analysis of how candidates feel about AI interviews.

First-time AI interview experience: every candidate rated it positive

Eight candidates in the dataset explicitly flagged that it was their first AI interview. Every single one rated the experience Good or Excellent.

Several described an initial moment of awkwardness that dissolved after the first few questions. One candidate put it plainly: the experience was different at first, a little strange, but after a few questions the conversation flowed naturally.

This pattern has real implications for rollout strategy. The fear is not that candidates try AI screening and reject it. The fear is that hiring teams never offer it because they assumed candidates would not like it. The data says the conversion from skeptic to supporter happens in about 90 seconds.

If you are holding off on AI phone screening because you think candidates are not ready, the candidates are ahead of you.

AI phone screens are less stressful than human interviews

This is the finding that surprises people most.

Multiple candidates reported that the AI format was less nerve-wracking than a traditional phone screen. The assumption is that talking to a machine adds anxiety. For a meaningful segment of candidates, the opposite is true.

Think about what a traditional phone screen feels like from the candidate side. You are being judged in real time by a stranger. There are awkward silences while the recruiter types notes. You can hear hesitation in their voice. You are rushing because they have 15 minutes between calls and you can feel the clock.

An AI screen removes most of those pressure dynamics. There is no interpersonal judgment happening in the moment. There is no rushed energy. Candidates can think through their answers without worrying about filling dead air.

For candidates who are anxious interviewers, particularly people who are strong performers but struggle with the social dynamics of a live screen, this matters. It means AI screening may actually surface better signal on candidate quality, not worse, because it removes a layer of noise that has nothing to do with whether someone can do the job.

That has real implications for candidate diversity too. Research from the Journal of Applied Psychology has consistently shown that interview anxiety disproportionately affects certain demographic groups. A format that reduces that anxiety is not just a better candidate experience. It is a more equitable one.

Multilingual AI interviews candidates did not expect

Seven Spanish-language positive responses appeared organically in the dataset. Candidates responded in Spanish with no prompting, across multiple organizations. Nobody configured those interviews to be bilingual. The candidates simply responded in the language they were most comfortable with, and the AI met them there.

For employers hiring in markets with large Spanish-speaking candidate pools, particularly in skilled trades, logistics, and manufacturing, this is not a nice-to-have. It is the difference between reaching your full candidate pool and missing a significant portion of it because your screening process only works in English.

Most human recruiters conducting phone screens are operating in one language. An AI screener that adapts to the candidate's language preference in real time expands reach without adding headcount. That is the kind of operational advantage that compounds over every open role. Our analysis of 70,000 voice AI interviews shows this pattern at scale.

Candidates call AI interviewing "the future of hiring"

21 candidates described their experience using words like "cool," "interesting," "new," or "the future of hiring," overwhelmingly in a positive context. This is worth paying attention to because it signals something beyond satisfaction. These candidates are not just tolerating AI screening. They see it as a step forward.

One candidate described it as the future of hiring. Another called it a great tool for streamlining and said they looked forward to seeing it improve as it got more data. That is the kind of feedback you get from someone who is not just satisfied but genuinely engaged with the format.

The narrative that candidates view AI as a step backward in the hiring experience is not supported here. For a significant portion of candidates, the framing is the opposite: this is where hiring is going, and they are fine with that. That tracks with what we see in the broader shift toward automated phone screening technology.

What AI interview feedback means for your screening process

If you are evaluating AI phone screening and candidate experience is a concern, here is what the data tells you.

The positive rate holds across industries. 83% positive is not a single-company result from a friendly audience. It spans staffing agencies, skilled trades employers, healthcare, logistics, distribution, automotive, and tech. The signal is consistent.

First-time AI interviewees are a conversion opportunity, not a risk group. Every single first-timer in this dataset rated the experience positively. The awkwardness resolves fast.

The "dehumanizing" concern is inverted. The most common unprompted theme in written feedback is that the AI felt human. Candidates are not experiencing what critics assume they are experiencing.

Anxiety reduction is a real benefit. For candidates who struggle with traditional phone screens, AI screening may produce better signal, not worse.

Multilingual capability expands reach without additional cost. Candidates will use it even when it is not explicitly offered.

What about negative AI interview reviews?

13% rated the experience Poor. Another 4% rated it Fair. That is nearly one in six respondents — and it is where the hot takes live. Not every candidate loved it. Those numbers are real and they deserve a serious look.

What really went wrong — and why the story is not what you think

We pulled every negative and mixed rating apart. Read every written comment. Categorized every root cause. The answer is not what most people expect, and it is the reason the adjusted positive rate climbs to 87% when you account for what actually went wrong (spoiler: a lot of it is not “AI is evil” — it is AirPods, bus Wi‑Fi, dead mics, and bad cell signal masquerading as a bad interview).

Part 2 goes line by line: every negative review, what caused it, and what that actually tells you about the gap between perceived and real candidate experience problems — including the difference between “this product failed me” and “my headphones failed me.”

We are also packaging the full methodology, dataset summary, and a stakeholder-ready quote bank as a downloadable report. That asset is still in production, and we will share it as soon as it is ready alongside Part 2.

How candidates actually feel about AI interviews

This post is Part 1 of a two-part series on AI interview candidate experience. Part 2 tackles the 13% Poor / 4% Fair cohort in depth — what broke, why, and what to do about it.