Expert warns AI hiring tools can cost companies the best candidates

Employers who let artificial intelligence judge job applicants risk losing the strongest candidates, IT expert and technology entrepreneur Yuri Gizatullin told Lenta.ru.

AI produces a polished report even when there is too little data about a person, and that assessment may not match reality at all. Gizatullin blamed recruiter laziness: it is easier to trust AI with an evaluation, since the report can be cited later if something goes wrong. But such a report can cost a company its best candidate, especially one with an unconventional career or an incomplete profile.

Without a clear system, AI may also start discriminating against applicants on various grounds. He cited an algorithm from American HR software developer Workday that screened out candidates over 40.

Nearly half of Russian companies already use AI in recruitment, but only about 12 percent apply it systematically across several hiring stages. In practice this turns into a box-ticking race. Gizatullin said the phrase "everyone is running, so I run too" captures the situation with AI adoption in business, HR included: a chatbot or AI feature is plugged into an HRM system, but the process itself is not rebuilt. The real problem is chaos in processes, which AI only accelerates. If a vacancy is described vaguely, selection criteria are not formalized and no one measures hiring quality, AI quickly and confidently produces bad decisions.

He called the war of algorithms the main problem. More than half of job seekers use neural networks in their search, and services already exist that find vacancies on hh and write cover letters themselves. As a result, the employer's AI reads resumes written by the candidate's AI, and no one can see a suitable candidate through that AI wall.

Gizatullin does not expect a return to manual search. He predicts the market will move to a hybrid model, where AI takes the routine and people make the final assessment. No workplace needs a "meat proxy" - a candidate who solves tasks with AI and passes the result off as their own. What is needed is a person capable of thinking and knowledgeable in the field. The structure of interviews will change, with more tasks performed live to rule out AI use. AI should take the recruiter's routine, but not the recruiter's brain. Those who silently screen and hire people by algorithm will face reputational damage and many unsuitable employees, he concluded.