I thought AI interviews were a bad joke. A 70,000-Person Study Proved Me Wrong.

A massive, 70,000-person field experiment just proved everything we thought we knew about AI job interviews is wrong. We see them as a dystopian joke—a lazy, inhumane way for businesses to get things done. I thought so too. But a new research paper tells a very different story.

Researchers put a conversational AI interviewer up against human recruiters in a real hiring process. The results weren’t just surprising; they were a complete knockout.

Here’s the data that should make every hiring manager and job seeker stop and listen:

• Better Hires: Applicants interviewed by AI were 12% more likely to get a job offer, 18% more likely to start the job, and 17% more likely to still be there after 30 days. The AI wasn’t just faster; it was better at picking winners.

• Candidates Prefer the AI: When given a choice, a staggering 78% of applicants chose to be interviewed by the AI over a human. Why? Less fear of being judged and more convenience.

• A Fairer Process: Switching to an AI interviewer nearly cut the number of reports of gender-based discrimination in half.

The bottom line is that a well-designed AI is a more consistent, thorough, and fair interviewer than a human. It doesn’t have bad days, get tired, or play favorites. It just gets the data.

This isn’t a story about robots taking over. It’s about a broken system for both candidates and companies, and how technology, against all our assumptions, might be the key to fixing it.

My awkward interview with Dave wasn’t an isolated incident. It was a symptom of a much larger disease. As professionals, solopreneurs, and creators, we operate in a hiring system that is fundamentally broken. We just don’t want to admit it.

For candidates, it’s a soul-crushing gauntlet. You spend hours customizing your resume and cover letter, only to send them into an application black hole, never to be seen again. You go through multiple rounds of interviews, getting your hopes up, only to be ghosted with no feedback. It feels cold, rude, and deeply inefficient. You start to believe it’s just a numbers game, a lottery where keywords and luck matter more than your actual skills.

But here’s the secret: it’s just as broken for the people doing the hiring.

Companies are drowning in applications. For example, 61% of HR professionals in Asia Pacific say it’s harder to find qualified talent, and many spend three to five hours a day just going through resumes. In high-turnover industries like customer service, companies are in a constant, expensive cycle of hiring. A bad hire costs a fortune, yet the tools they use—gut feelings and biased first impressions—are terrible at predicting success.

This creates a toxic feedback loop. Companies, overwhelmed by volume, use flawed shortcuts to screen people out. Candidates, knowing the odds are against them, resort to a “spray and pray” method, flooding the system even more. The result? Good candidates are overlooked, bad hires are made, and everyone ends up frustrated.

The “hustle culture” advice of sending out 100 applications a day isn’t a strategy for success. It’s a symptom of a broken system. More applications isn’t the answer. A better system is.

For years, my solution to the hiring problem was simple: avoid it. As a solopreneur, I focused on building a business of one. But the problem was always in the back of my mind. A few weeks ago, I found a research paper that completely changed my mind.

It wasn’t some theoretical nonsense from an ivory tower. It was a huge natural field experiment called “Voice AI in Firms.” Researchers worked with a global recruitment firm to look at more than 70,000 real job applications. They randomly chose applicants to be interviewed by either a human recruiter or an AI voice assistant named Anna.

My first thought? Yikes. A dystopian nightmare. The ultimate dehumanization of the hiring process.

I was dead wrong. The results weren’t just surprising; they were revolutionary.

Let’s get right to the data, because it’s staggering. The scientists tracked what happened after the interviews, and the AI was better than humans in every important business metric.

Applicants interviewed by the AI were:

• 12% more likely to get a job offer.

• 18% more likely to actually start the job.

• 17% more likely to still be working after 30 days.

Read that again. The AI wasn’t just a faster, cheaper option. It was more effective. It helped the company find candidates who were not only a good fit on paper but also succeeded and stayed in the role. The best part? When the researchers surveyed the company’s own human recruiters beforehand, most of them predicted the AI would do worse.

This was the finding that truly blew my mind. We assume people crave human interaction and that talking to a bot would be a huge turn-off. The data says the exact opposite.

When applicants could choose between a human and an AI interviewer, a shocking 78% chose the AI.

This wasn’t a fluke. The preference was consistent. So, why would anyone choose a robot over a person? The study points to a few powerful reasons. First, convenience. The AI is available 24/7, eliminating the frustrating back-and-forth of scheduling. Second, and more profoundly, it seems to reduce the fear of judgment. The study found that applicants with lower standardized test scores were far more likely to choose the AI. This suggests that for many, the stress of being judged by a human—on their accent, their nervousness, or their “vibe”—is a greater burden than the interview itself.

And the experience was just as good. Applicants rated the AI interview and the “recruiter” on par with the ones led by people. The Net Promoter Score, which shows how likely you are to recommend a company to a friend, was almost the same for both groups.

The AI is a Better Interviewer (And It’s Not Even Close).

So, how is this possible? How can an AI outperform a person at such a quintessentially human task?

The answer is consistency.

The researchers analyzed thousands of interview transcripts and found that the AI was simply more thorough. Human recruiters were all over the place, maybe because they were tired or biased. The AI, on the other hand, was a model of consistency.

• It was more comprehensive: AI-led interviews covered significantly more of the required topics than human-led ones.

• It got better information: The AI’s structured approach made applicants give more hiring-relevant details. Transcript analysis showed that candidates in AI interviews used more of the specific language features—like a richer vocabulary and more complex sentences—that are statistically proven to predict who gets a job offer.

The AI doesn’t have bad days. It doesn’t get tired. It doesn’t play favorites. It gives every candidate the same fair and thorough evaluation. Every time, it follows the best-practice script perfectly.

The AI is a Fairer Interviewer.

This might be the most important finding of all. The traditional interview is a minefield of unconscious bias. We are naturally drawn to people who look, sound, and think like us. The study showed a powerful way to fight this.

Changing from a human to an AI interviewer cut the reported rate of gender-based discrimination nearly in half, from 5.98% to 3.30%.

By taking the human out of that first, high-stakes interaction, you remove a primary way for bias to creep in. This isn’t just about making business more efficient; it’s about making hiring processes fairer and more just. The data shows that for many candidates, especially those from underrepresented groups, the fear of human judgment is a real and significant barrier. An AI, seen as an impartial system judging their words instead of their identity, creates a safer space. It levels the playing field, letting the quality of their answers—the substance—shine through, instead of being filtered through an interviewer’s personal style or biases.

So, if the data is this clear, why does the idea of an AI interview still make us shudder?

This is where things get complicated. The study’s groundbreaking results are a stark contrast to the overwhelmingly negative feelings you find in the real world. There’s a big difference between what this technology can be and what it often is.

The View from the Trenches: “Dehumanizing” or “Efficient”?

Spend five minutes on Reddit, and you’ll see the candidate’s point of view in its raw, unfiltered form. People say AI interviews are “dehumanizing,” “lazy,” and an instant “red flag.” Many feel it ruins the “two-way street” nature of an interview, turning it into a cold, one-sided evaluation where they can’t ask questions or build rapport.

But then, look at it from the company’s side. Go to a software review site like G2, and you’ll see HR managers praising these tools for their efficiency. They love being able to screen hundreds of candidates in the time it used to take to screen a dozen, cutting down on hiring time and administrative costs. For companies that get a flood of applications, automation feels less like a choice and more like a necessity. In fact, 74% of hiring professionals hope AI will handle these boring tasks so they can focus on more strategic work.

The problem often lies with the technology itself. The AI in the study was an advanced, interactive conversational agent that had a real-time, back-and-forth dialogue. Many of the commercial tools that have shaped public perception are much simpler: they are “one-way” video platforms where you record yourself answering pre-set questions. This experience feels less like an interview and more like an audition tape, which understandably feels impersonal.

Cheating the System: When AI Interviews AI.

This disconnect has created a bizarre technological arms race.

As companies have adopted AI to screen candidates, candidates have started using AI to pass the screening. Recruiters say that more than half of the people in technical interviews are now using AI tools in real time to solve coding problems and answer questions. They repeat the question to feed it into a tool like ChatGPT and then read the generated answer off a second screen.

The corporate response has been swift. To make sure candidates have the basic skills they claim and to fight this new wave of AI-assisted cheating, tech giants like Google and Amazon are bringing back mandatory in-person interview rounds.

This isn’t just a simple case of cheating. It’s a systemic failure. When the hiring process feels like a game—where you stuff your resume with keywords to beat an Applicant Tracking System (ATS)—candidates will use the best tools they can find to win. This creates a cycle of distrust that is bad for everyone and makes it hard to tell who is really qualified.

This new world can be scary. But it’s also a huge opportunity. The rules are changing, and those who know the new rules will have a big advantage. The study gives us a clear path.

For Job Seekers: Stop Fearing the Bot, Start Acing the Interview.

The AI interview isn’t an obstacle to be feared; it’s a chance to be judged on your ideas, not how smoothly you speak. Here’s how to nail it:

• Treat it like a person (kind of). The AI has learned from millions of human conversations. It expects natural language. Speak clearly and at a normal pace. Don’t try to sound like a robot or fill your answers with a bunch of random keywords. The goal is clarity, not code.

• Focus on Substance, Not Style. This is your time to shine. Unlike a human who might be swayed by charisma, the AI is analyzing the content of your answers. The study found that it rewards complete answers, a rich vocabulary, and specific examples. Don’t worry about being nervous; worry about being thorough.

• Use the Job Description as Your Cheat Sheet. The AI has a clear set of rules based on the job description. Before the interview, read it line by line. For each skill or requirement, prepare a specific story or example from your own experience that shows you are competent. Naturally weave those words and ideas into your answers.

• Forget the cheating tools. The arms race is a waste of time. You will get caught, and it shows a lack of integrity. Instead, use AI as a preparation partner. Use tools like ChatGPT to make practice questions, improve your stories with the STAR method, and get feedback on your clarity and conciseness before the real interview.

For Founders & Hiring Managers: Your New Unfair Advantage.

For anyone building a team, this is a chance to leapfrog the competition by building a smarter, fairer, and more effective hiring machine.

• Embrace the “Human + Machine” Model. This is the most important lesson from the study. Don’t try to automate hiring completely. Use AI for what it does best: screening a lot of people at the top of the funnel in a consistent and unbiased way. Then, use your human recruiters for what they do best: building rapport, assessing cultural fit in the final rounds, and selling your top candidate on your company’s vision.

• Choose the Right Tool. Not all AI is the same. Avoid the old, clunky one-way video tools that alienate candidates. Invest in modern conversational AI platforms that provide a seamless and positive experience. Remember, the hiring process is the first impression a candidate has of your brand; make it a good one.

• Be Radically Transparent. The AI in the study quickly identified itself as an AI and explained that a person would make the final decision. Do the same. This simple act builds trust and manages expectations, turning a potentially jarring experience into an understandable part of the process.

• Rethink Your Signals of Quality. The study found something interesting: when human recruiters evaluated candidates interviewed by the AI, they made the right call. They put less weight on the subjective interview score and more on objective data like standardized test scores. This is the way to go. Trust the AI to get consistent data, but use multiple data points—like an interview, a work sample, or a skills test—to make your final, holistic decision.

The researchers in the study built a custom AI from scratch. The good news is that you don’t have to. There are now new commercial platforms that can give you many of the same benefits, whether you’re a founder hiring your tenth employee or a professional getting ready for your next big move.

The All-in-One Platforms (For High-Volume Screening)

These are for businesses that need to screen a lot of candidates quickly and consistently.

• HireVue is the clear leader in this space. It has a lot of powerful tools, such as video interviews on demand, game-based psychometric tests, and AI-driven scoring. It’s very strong, but it costs a lot—usually starting at $35,000 a year—so it’s best for big companies.

• Paradox (Olivia): This platform is built around its conversational AI assistant, Olivia. She talks to candidates 24/7 through SMS and web chat, doing initial screening, answering questions, and setting up interviews on her own. It’s a game-changer for hourly jobs with a lot of applicants in fields like retail and hospitality.

• HeyMilo: A strong contender focused on high-volume hiring for industries like BPOs. It has AI-powered voice and video interviews, resume screening, and great support for multiple languages, so it can interview candidates in their native language.

• Manatal is a great, low-cost choice for small to medium-sized businesses and startups. It has a full Applicant Tracking System (ATS) and an AI engine that suggests the best candidates from your talent pool and adds public data to their profiles.

The Interview Co-Pilots (For Augmenting Human Interviewers)

These tools aren’t meant to replace people; they’re meant to make them better, more focused, and less biased.

• (https://brighthire.com/): This is an “interview intelligence” platform. It works with your video calls (like Zoom and Teams), records and transcribes them, and then uses AI to pick out the most important parts and make notes automatically. This lets the interviewer be fully present in the conversation instead of typing like crazy, which leads to better decisions.

• Metaview: Like BrightHire, Metaview is all about automating the boring task of taking notes. It joins your interviews, makes structured notes that match your hiring criteria, and even writes feedback summaries, cutting down on the time you need to spend on paperwork after the interview to just a few minutes.

The Candidate Prep Tools (For Leveling the Playing Field)

For job seekers, these tools are the ethical and effective way to use AI to your advantage.

• Interviews.chat / Interview AI: These platforms act as your personal, on-demand interview coach. You can upload a job description, and the AI will make questions that are specific to that job for you to practice with. It gives you feedback on your answers in real time, which helps you improve your stories and gain confidence before you even talk to a real recruiter.

• (https://www.finalroundai.com/): This tool has a lot of prep features, like an AI Resume Builder and mock interviews. It also has an “Interview Copilot” for real-time help, which is part of the “arms race” but can be a great way to practice how to organize your thoughts when you’re under pressure.

To make it easier to navigate, here’s a quick breakdown of the top tools:

The fear surrounding AI is that it’s here to replace us. The data tells a different story. The future isn’t AI versus human; it’s AI plus human.

The study on voice AI in hiring provides the perfect blueprint. The most effective, efficient, and fair system used a machine for what machines do best: processing a lot of data consistently and without bias to handle the top-of-funnel screening. This freed up people to do what they do best: making nuanced judgments, building real relationships, and making the final strategic decision.

This isn’t replacement. It’s elevation. It’s about automating the boring, soul-crushing parts of our jobs so we can focus on the creative, strategic, and deeply human work that really matters. As platforms like LinkedIn roll out their own AI-powered recruiting tools, this is the main message: it’s about making recruiters more effective, not obsolete.

The world of work is changing faster than ever. You can either be disrupted by it or you can lead the change. The tools are here. The data is clear.

The only question is what you’ll build with them.