AI interviewers promise unbiased hiring
“You met someone you like at a blind date and took her to an upscale restaurant. You told her that you will pay for the meal but realize that you left your wallet at home. How will you explain the situation to your date?”
This is an actual scenario question that job candidates will have to answer in less than two minutes if they go into an interview powered by “inAir,” an AI-powered recruitment system. InAir is the brainchild of software developer Midas IT, which presented the system on March 7 at The-K Hotel in Seocho District, southern Seoul.
Candidates taking an interview with inAir wear headsets equipped with a mic and register their face and voice with the system before beginning the interview. They then state their strengths and weaknesses, answer a series of situational questions and match words describing emotions to images of individuals with different expressions. The candidates also play a simple online game.
The inAir system analyzes a candidate’s expressions and muscle movements during interviews. Nothing escapes the scrutiny of the AI interviewer, which keeps track of everything from the candidates’ tonality and vocabulary choice to their heartbeat.
“Businesses can assess applicants on various fronts using our system, grounded in neuroscience and biology,” explained Midas IT CEO Lee Hyung-woo.
In January, SK Hynix applicants had their written applications assessed by sister company SK C&C’s AI computing platform Aibril. Lotte Group announced last month that it would also use AI technology to screen documents during this spring recruitment season.
One of the biggest perks that AI recruiters bring to the table is the removal of human bias. While there’s no room for subjectivity in machine-conducted interviews, human recruiters may let their personal taste, preconceptions and mood subconsciously influence how they assess candidates.
“AI recruiters are getting more attention as unfair hiring practices continue surfacing,” one HR specialist said. “We expect the machines to help allay concerns over current recruiting practices.”
The AI recruitment systems are also useful for making the hiring process more time- and cost-efficient. Machine recruiters take an average of three seconds to assess an autobiographical essay, or eight hours to grade 10,000 of these essays. It would take 10 human recruiters 56 hours to complete the same workload.
Other countries are also using smart machines to acquire talent. Last May, Japanese tech conglomerate Softbank adopted a hiring solutions platform powered by IBM’s Watson to review job applications. The program, after learning and analyzing Softbank’s past recruitment data, screens documents to single out candidates who display characteristics preferred by the company.
IBM itself uses AI technology to interview candidates by voice or video call and then invites a select few to participate in an interview of greater depth with a human recruiter.
Critics warn against depending wholly on machines to hire talent, however. One reason is that computers, unlike humans, only rely on the data they can see, and thus cannot evaluate candidates on their future growth potential. Some applicants interviewed by AI recruiters have also complained that they couldn’t get used to talking to computers, and that the experience felt uncomfortable and unnatural. Recognizing these deficiencies, most companies using AI-powered hiring methods are still keeping human recruiters.
“Humans are best suited to judge a candidate’s character and potential,” said Seo Mi-young, CEO of job portal site Incruit. “We can use the objectivity and efficiency of the AI recruitment system, but only as a reference in making the final decision.”
BY CHOI HYUN-JU [email@example.com]