AI Interviewers Are Triggering the “Uncanny Valley” Effect, Chinese Job Seekers Say

October 3 (IT Home) — The hashtag “AI interview uncanny valley” trended today (3 October). During the recent autumn recruitment season, a large number of companies have brought in AI digital-human interviewers, whose stiff expressions and panoramic micro-expression monitoring reduce interviews to cold algorithmic scoring, prompting job seekers to complain en masse about the “uncanny valley effect.”

By the time IT Home published this article, the topic ranked 25th on the trending list, with a related heat value of 186,858. In July this year, the “AI face uncanny valley effect” had also trended on Weibo, mainly in connection with AI short dramas.

The “uncanny valley effect” refers to a psychological phenomenon in which, when a non-human entity — such as a robot, a virtual character, an AI voice or a digital human — is very close to a human in appearance, voice or behaviour, yet has subtle unnatural qualities, people’s feelings suddenly shift from “liking” to “strong discomfort or even fear.”

According to feedback from Weibo users, AI digital-human interviewers do blink and pause for breath when speaking, but their gaze is empty and their micro-expressions are stiff, producing a “human-like yet not human” plastic quality that directly triggers perceptual conflict and physical revulsion in the human brain.

The system uses a camera to track pupil movements and facial muscle tension in real time, converting unconscious physiological reactions — such as delayed answers or a trembling speaking pace — into concrete scores. Many users feel that AI interviews are disrespectful to people.

A Weibo user who posts about cars shared a video of an AI interview and said: “Looking for a job now, nine times out of ten you first have to get past the AI stage: answering questions to the camera while the system automatically analyses your expressions, speaking pace and keywords. A real HR person may never appear from beginning to end.”