For most of my career, I trusted the interview. Not because I thought it was scientifically sound, but because of how it made me feel. A good interview left me confident. I liked the way the candidate carried themselves, how quickly they answered a tough question, how the conversation flowed. I walked away thinking, "I've got a good read on this person”. 

I always suspected that feeling wasn't the same thing as evidence. I just didn't have a reason to question it, because it was the process everyone used, and it gave the hiring side a sense of control. The truth is that the interview was never really measuring whether someone could do the job. It was measuring how someone performed under a very specific, very artificial kind of pressure, and we mistook our own comfort with that performance for a good signal.‍

That gap has always existed. Remote work and AI just made it impossible to keep ignoring.

‍We Built Confidence, Not Validity

The research on this isn't new, and it isn't kind. Schmidt and Hunter's landmark meta-analysis found that unstructured interviews, the free-flowing, conversational kind most of us grew up doing, only lead to the right hire slightly more often than a coin flip. A more recent 2022 re-analysis by Sackett and colleagues puts the predictive validity of unstructured interviews even lower. Structured interviews do meaningfully better. A 2025 meta-analysis led by Wingate, drawing on 37 studies and more than 30,000 participants, found that structured interviews predict task performance at roughly twice the rate of unstructured ones.

‍What we were actually good at assessing was composure, quick thinking, and social ease in a stressful, unfamiliar room. What we told ourselves we were assessing was job performance.  Those are not the same skill, and for a lot of roles, they barely correlate.

‍Remote Work Removed the Noise We Mistook for Signal

In person, we had all sorts of extra cues to lean on: the handshake, the small talk in the lobby, the way someone carried themselves walking to the boardroom. None of it was predictive of job performance either, but it added texture to our gut feel and made the process feel more trustworthy. Remote hiring stripped a lot of that away. Suddenly it was just a face on a screen, answering questions, with none of the ambient context we used to lean on. As Kristina Benassi, Senior Manager of Talent Acquisition at Euna Solutions, put it on a recent episode of the Unmuted Podcast, remote hiring optimized for convenience, not confidence. We didn't lose a good process. We lost the noise that made a mediocre one feel reliable.‍

‍Then AI Made the Cracks Impossible to Ignore

Once the noise was gone, AI walked straight through the opening. A Gartner survey of more than 3,000 job seekers found 6% admitted to some form of interview fraud, whether impersonating someone else or having someone impersonate them, and Gartner projects that by 2028, one in four candidate profiles globally will be fake.

The employer side tells the same story. A 2025 Checkr survey of 3,000 managers found 59% had personally suspected a candidate of using AI to misrepresent themselves during the hiring process, and a 2025 GetReal Security survey of risk and fraud leaders found 41% confirmed their organization had actually hired and onboarded a fraudulent candidate. Pindrop's 2025 Voice Intelligence and Security Report found deepfake fraud attempts jumped more than 1,300% year over year between 2023 and 2024.

Those numbers aren't a story about a few bad actors gaming the system. They're a story about a hiring process that was already running on a weak signal, and technology just made the weakness visible.‍

‍More Interviews Isn't a Fix. It's More of the Same Flawed Signal.

The instinct, understandably, is to add friction. More rounds. More assessments. More verification steps. Kristina made a point on the podcast that stuck with me: piling on more of the same kind of evaluation doesn't necessarily produce a better hiring decision. If a 45-minute conversation has weak predictive validity, conducting five 45-minute conversations don't fix that math. They just take longer to arrive at the same weak signal. The better use of that effort is redesigning what you're actually measuring, not multiplying the number of times you measure it.

‍What Trust-Built Hiring Actually Looks Like

Kristina drew a distinction I think more HR and Talent leaders need to sit with: the difference between having a hiring process and having a hiring strategy. A process is a sequence of steps. A strategy is a deliberate choice about what evidence actually predicts success in the role, then building the steps around that.

In practice, that looks less like a gut check and more like:

• Structured scorecards with the same core questions for every candidate, scored against defined criteria
• Work samples and job-relevant exercises instead of open-ended conversation, especially for technical and senior roles
• Reference checks that ask about specific past behaviour, not general impressions
• Interview panels trained to separate "I liked them" from "the evidence says they can do this"

None of this eliminates human judgment. It just stops asking judgment to do a job it was never built for.

‍A Final Thought

I still remember the interviews that made me feel great about a hire, and a few of them didn't work out. I remember the interviews that felt a little flat, where the person turned out to be one of the best hires I ever made. The feeling was never the data. It just felt like data because it was familiar.‍

Remote work and AI didn't break hiring. They took away the comfortable noise that let a weak signal pass as a strong one. The organizations that figure this out next won't be the ones running more interviews. They'll be the ones finally asking what their interviews were actually measuring all along.

If you want to watch/listen the full conversation with Kristina Benassi on designing trustworthy, authentic hiring in the age of AI and remote work, check out episode 82 of the Unmuted Podcast, powered by ScaleHR.

‍