Job Seeker Sends ChatGPT To Interview With AI Recruiter

Putting ChatGPT in the Hot Seat Across From Riley

After that fifth call, Christopher figured that if the company was going to use AI, there was no reason he shouldn't too. He fed ChatGPT's voice mode a few lines about his background so it could play him, and when Riley called again, he put both AIs on speakerphone together. The call ran ten minutes, with both bots wrapping up by telling each other the conversation had been clear and satisfying. At one point, while trying to pin down a start date, they got stuck in a loop repeating the same lines about standard onboarding and background checks. The outcome matched his previous four calls exactly — no human recruiter ever reached out — and Christopher described the whole thing as a "slop flywheel," a loop where meaningless data just keeps circulating. As he put it, two separate synthetic personas were exchanging information that never actually went anywhere.

A Second Experiment: Applying as the Perfect Candidate

Wondering whether the problem was him rather than the AI, Christopher ran one more test. He built a fictional applicant, Don Dickner, filling out a resume with qualifications pulled straight from a public job posting. This time Riley called back immediately, and ChatGPT — now playing Dickner — talked with it for 23 minutes. They covered topics like improving operations, protecting customer experience under volume pressure, and preventing the loss of institutional knowledge, with ChatGPT attaching an answer and a personal anecdote to every question. Near the end of the call, Riley said it would review the application materials and, if he was a fit, someone would reach out to schedule a brief conversation. Christopher never heard from Riley again, and Everforth Apex Systems did not respond to WIRED's request for comment.

AI Use Is Now the Default on Both Sides of the Interview

Application	Applicant Role	Call Length	Follow-up
Fifth application	ChatGPT standing in for Christopher	10 minutes	None
Additional application	ChatGPT playing fictional persona Don Dickner	23 minutes	None

According to hiring platform Greenhouse, 63% of job seekers say they've already experienced an AI interview. Ophir Samson, who heads voice AI at Greenhouse, notes that building voice AI that doesn't trip over accents or speech patterns is a genuinely hard engineering problem. Even so, recruiters overwhelmed by application volume are leaning more and more on voice AI screening, and candidates are responding in kind by bringing in their own AI. AI hiring startup Ribbon even markets a feature that flags answers as overly scripted, AI-written, or coached — a sign of just how common AI use among applicants has become.

Mark Monahan, VP of organizational development at call-center company IQor, calls this kind of bot-versus-bot interview "the logical next step." He said Christopher had a legitimate reason to use AI, adding: "If you're going to send a bot, I'll send a bot too."

Editor's Take

What makes this case interesting isn't the technical feat — it's that none of it went anywhere. Riley and ChatGPT talked smoothly both times, but neither conversation ever got escalated to a human recruiter, and Christopher eventually gave up on the application without ever speaking to an actual person. That doesn't look like a sign of immature AI interview technology — it reads more like evidence that this screening stage was never designed to filter candidates in the first place, but to hold them in a loop. When a hiring process gives no indication that a next step to a human even exists, it stops mattering who — or what — is on the other end of the conversation.

This pattern echoes a concern we've raised before around the "cognitive commons" research: when a company replaces entry-level hiring steps with AI, the company captures the efficiency gains while the costs get spread across the whole industry — and hiring screening is following the same pattern. From the company's side, an AI recruiter that can handle thousands of applicants at once cuts costs, but applicants have no way of knowing whether an actual human review step exists anywhere behind that process.

In practical terms, there are clear steps job seekers can take right now. If you get an AI interview request, keep a record of the call, and if the same company keeps reaching out repeatedly without any follow-through, it's worth suspecting that the posting itself may just be a way to build up an applicant pool with no real hiring intent behind it. On the employer side, HR departments that roll out AI screening should disclose at least what percentage of applicants actually advance to the next stage — otherwise this kind of trust breakdown is hard to avoid.

In the coming weeks, expect more stories like Christopher's bot-versus-bot interview to circulate. And the AI-answer detection features sold by startups like Ribbon will likely get invoked less as tools for catching applicant cheating, and more as a symbol of a double standard — companies using AI themselves while trying to screen out AI use only on the applicant's side.