What happens when a job interview leaves the script behind? One new video shows an AI interviewer trying to process a candidate who answers with raccoon protocols, plumes of gauze, and a pharmacist-shaped hallway.
The result is part comedy, part stress test, and part warning about automated hiring. As AI interviews reach more candidates, these strange exchanges reveal how fast a polished system can lose its footing when a conversation moves beyond familiar patterns.
One Candidate Sends an AI Recruiter Into Confusion
Graham Zip, a satirical influencer, launched a faux-hallucinated spiel during a job interview with Dana Whitfield, an AI avatar. Instead of offering clear qualifications, Graham introduced a fictional identity and a stream of disconnected phrases that pushed the virtual recruiter outside its comfort zone.
“I’m Chick Haversplat, or, Chick Bongo for short. And uh, Q3 pipeline, Salesforce, Kayak margin,” Graham said. The AI bot answered, “It’s a pleasure to meet you, Chick Bongo,” accepting the invented name without questioning the sudden shift in the conversation.
Graham continued with another strange credential: “Six years of sprucewood — actually seven, if you count Richard.” Dana Whitfield translated the statement into a conventional employment history, replying, “It sounds like you built up a significant tenure over the last seven years, including your time with Richard at sprucewood.”
The exchange kept bending away from ordinary interview language. Graham described his experience as “Mostly palette forward, a little bit of denim,” then added, “You know, every pharmacist is a hallway.” Dana responded that Graham had “a unique way of navigating the palette-forward and denim aspects” of his work.
Then came the phrases that pushed the interview toward full breakdown. Graham explained that he was looking for a job with “two plumes of gauze” and to follow “the raccoon protocols.” Dana tried to make sense of it, saying, “It sounds like you’re looking for a position where you can really lean into those raccoon protocols, and the plumes of gauze.”
By the end, the AI bot began producing its own nonsensical statement: “There was a bit of a gap in the drum counting process, after Tuesday, with the work truck and cousin Francis.” Yet the interview still closed with a familiar promise: “It has been a pleasure chatting today Chick Bongo, and I will be in touch very soon about the next steps.”
It isn’t clear how Graham’s application progressed or what role the interview was for, but the AI’s enthusiasm suggests things went well.
A Bot Interviews Another Bot
Another case turns the hiring process into a closed loop. Christopher, a job seeker, applied to roughly 700 jobs over six months, and five applications went through Riley, an AI recruiter for Everforth Apex Systems. After receiving no follow-up or rejection, he used ChatGPT Voice to simulate a job interview with Riley.
Christopher created a fake candidate named Don Dickner, using qualifications from a job posting. Riley reached out after the application submission, and ChatGPT spoke as Dickner during the interview. Christopher described the result as a “slop flywheel,” mixing mischief with frustration.
The setup produced a bot-on-bot interview that lasted 23 minutes. ChatGPT, speaking as Dickner, discussed various professional topics with Riley, while another ChatGPT simulation involving onboarding and background checks lasted 10 minutes and circled through the same points.
At one point, ChatGPT responded as Dickner: “I appreciate it. Upon carefully reviewing your application, we will assess your qualifications for the position. Should you meet our criteria, a recruiter from Everforth Apex will reach out to schedule a brief conversation.” The automated exchange mirrored the language of hiring systems while sending no human into the conversation.
“I just put them next to each other, and they just hit it off,” Christopher said. Then the connection ended. Christopher never heard from Riley again after the bot-on-bot interview, and Everforth Apex Systems did not respond to a request for comment.
AI Interviews Are Already Part of the Hiring Funnel
These episodes land as companies add voice AI to job interviews, even though the technology faces problems with accents and interruptions. A ManpowerGroup survey found that 63 percent of job seekers report they’ve encountered an AI interview.
The pressure now runs in both directions. Candidates are using AI to assist in interviews, while startups like Ribbon offer to identify AI-assisted responses. Employers are building systems to evaluate candidates, and candidates are using systems to test, support, or replace their own answers.
Mark Monaghan, vice president of organizational development at IQor, and Ophir Samson, head of voice AI at Greenhouse, are among the names connected to the wider discussion around AI in hiring. Kyle Elliott, a Forbes career coach, is also listed among those involved in the conversation.
The question is no longer whether AI will appear in job interviews. It already has. The sharper question is whether automated recruiters can recognize a genuine answer, handle an interruption, and stop a circular exchange before it turns into another “slop flywheel.”
Rumors that bot-on-bot interviews could preview future hiring practices give these strange moments a bigger edge. If recruiters and candidates both delegate conversations to AI, the hiring process may produce polished language without producing a clear human decision.
For now, the lesson is visible in both interviews: AI can keep a conversation moving even after meaning starts to disappear. That ability may help systems manage ordinary applications, but it also makes the failures harder to spot. The next stage of AI hiring will depend on whether companies treat these breakdowns as jokes, glitches, or signs that a human needs to step back into the room.
Based on
