"Is This Applicant a Real Person?" Is Now a Screening Question

In the Fortune piece on hiring's "vicious cycle," Indeed CEO Hisayuki Idekoba framed the first problem in a way that would have sounded strange a few years ago: "How can we make sure job seekers are a real person, and it's not an AI application, and they're serious, and they're qualified."

That's four separate questions: real, not AI-generated, serious, qualified. A resume answers none of them reliably anymore. It's worth being precise about which ones a screening step can help with, because vendors are going to promise all four.

What's changed about the applicant pile

The Fortune article describes applicant pools flooded with AI-generated applications, and says scams involving fake candidates have made it harder for companies to identify legitimate applicants. Even setting the outright fakes aside, the everyday version is simpler: a polished resume and cover letter now cost a candidate almost nothing to produce, so polish no longer tells you anything about effort, seriousness, or skill.

What we can tell you, and what we can't

We'd rather be direct about the limits. TrueScreenHR is not an identity-verification service. We don't check a government ID or confirm a degree with the school that issued it. If identity fraud is your core risk, you need a tool built for that, and job boards are adding verification on their side too.

What a structured interview can do is test the other three questions — AI-generated, serious, qualified — in a way a resume can't:

• Serious: an invited applicant has to sit through 17 interview questions and answer them in writing. People mass-applying to hundreds of jobs rarely finish that for a role they don't care about. Who completes the interview is itself signal.
• AI-assisted: the interview records paste events, how much of each answer was pasted rather than typed, and answers submitted suspiciously fast. Those feed an integrity score and a fraud-risk rating you see next to every candidate.
• Qualified: answers are cross-checked against the candidate's own resume, title, and years of experience. Claiming senior-level ownership the answers don't back up gets flagged, separately from the behavioral signals.

None of these signals is proof on its own, and we don't present them that way. A high fraud-risk rating is a reason to look closer at the transcript and ask follow-up questions in a live interview, not an automatic rejection. But it's a far better starting point than a resume that looks exactly like the other 400.

A screening step can't prove who someone is. It can show you whether they did the work themselves, whether they cared enough to finish, and whether their answers match their resume.

If AI-written applications are making your pile harder to trust, you can invite your current applicants to the interview and see the integrity read for yourself. The first five invites are free.