What if your first job interview says, “Tell me about yourself,” but there is no person on the other side?
That is not a strange idea anymore. In 2026, Greenhouse said its Ezra AI Labs deal would bring voice and conversational AI into hiring, and Ezra’s system can run structured 15-minute voice interviews that score skills, communication, problem-solving, and possible fraud signals. Greenhouse also says AI interviews are now common: in its 2026 report, 63% of U.S. job seekers said they had already experienced one. (greenhouse.com)
Now picture Mika, sitting at her kitchen table after dinner. She clicks a link. A calm AI voice asks the first question. Mika answers, then starts to worry. Should she sound more cheerful? Did that pause hurt her? Is the system judging her ideas, or her voice, or even her accent?
And here is the turn: many applicants are not simply afraid of AI. They are confused by unclear rules. Greenhouse says candidates are pushing back more on unclear and inconsistent use than on AI itself. In the same report, 70% of U.S. candidates said AI was not clearly disclosed before their most recent AI interview, and only 21% believed employers were using AI responsibly and transparently. Gartner found that only 26% of job candidates trust AI to evaluate them fairly, and 32% worry AI could fail their application. (greenhouse.com)
There is another reason for that worry. U.S. disability guidance says voice or facial analysis can screen out qualified people with speech or other disabilities unless employers test the tools carefully and offer accommodations. In Illinois, the law says employers using AI analysis in recorded video interviews must tell applicants first, explain in general how the AI works, get consent, and delete videos within 30 days if the applicant asks. A 2026 study found the same pattern again: unclear evaluation rules, weak support, and a feeling of distance made applicants more stressed. (ada.gov)
So maybe the biggest problem is not the machine asking the question. It is the silence around what the machine is really doing. And when people do not know the rules, even a simple question can feel very hard.










