The interview before bed
A candidate for a customer-service role opened an AI interview before bed. It was their first, and they treated the automated screen casually.
They expected the job-specific questions to come later, with a person. “I thought those questions would be reserved for a human interview, but I was wrong.”
Instead, those questions arrived inside the AI interview. What looked like a first step became the assessment they had not known to prepare for. When it ended, the candidate was left wondering whether the interview had gone well—or whether they had already damaged their chance at the role.
The failure in this account is not that the interview used AI. It is that the candidate did not know what kind of assessment they were entering.
They expected one sequence: an automated first step, then job-specific questions with a person. The process used another. By the time the mismatch became visible, the assessment was already under way.
The account ends without an outcome. The candidate does not know whether the interview went well or whether they had damaged their chance at the role.