Artificial intelligence is now entering the field of HR and is changing the very basis of working with people. According to Gitnux, already 70% of companies in the world use AI at different stages of mass placement, and 58% use it for automatic screening of resumes. And the Workable report says: companies that have implemented AI in recruitment increased their hiring efficiency by 89%, and the closing rate of vacancies has increased by 44%. HHilton, for example, were able to reduce hiring time by almost 90%, automating standard steps. While in India 75% of employers allocate up to 70% HR-budget on AI-tools.

The logical question arises: can artificial intelligence ever completely replace a live recruiter? Alexander Fedulov, head of the SENSE selection group, shares his practical experience, discusses the capabilities and boundaries of neural networks and explains why recruitment is primarily about people.

What does the recruiter do and what is already being automated

Recruitment is not about «send 20 resumes and put interviews in free windows». Real work takes place deeper: the specialist understands the motivation of the candidate, sees how he will reveal himself in the team and where he will not grow.

It’s always about people: about trust, honesty and subtle interaction. However, technology has already irrevocably changed our hiring landscape. What is already working in recruitment:

• AI is able to write accompanying letters not on the template, but adapting to the specific job;
• Algorithms handle the primary screening of resumes, especially key words and basic filters;
• ML-models, which analyze video interviews and identify key behavioral markers according to given criteria;
• Systems that automatically collect short profiles of candidates, sort them by relevance and even conduct primary dialogue through bots.

All these tools are not about «replacement», but about unloading. Already now 25% of tasks can be automated with AI, which will help to free the recruiter from routine and give space for a deeper interaction with people.

Loss of automation: where does the recruiting soul go?

The introduction of AI in personnel selection is definitely a step forward. The program can filter hundreds of resumes in a few minutes, highlight skills and even evaluate tone. The speed of processing all information about the candidate becomes higher, which allows to focus on strategic aspects of the work.

But there is one «but»: all this works only according to the given patterns, and read between the lines of the algorithm until you have learned. Artificial intelligence can not feel how a person really «own», see motivation, true character, some nuances of the behavior of the candidate, for example, uncertainty or glare in the eyes when talking about the future. All these nuances are often revealed only in a live communication.

One of the most striking cases in my practice is the selection of a person for my team. At the interview came a strong candidate: confident skills, relevant experience, clear pitch.

I wanted to not just analyze, but пропустить этот опыт через себя, команду и ценности, on which it is held.

And theni got the point: when he talked about work, it all came down to a set of conditions: «here is my area, there are borders, the rest – not mine». that’s not how we work.

We are building a system where the real participation is important, not «role play». I hesitated for a long time, because from the rational point of view it was perfect, but inside everything protested: if I say «yes», it would be a cold, empty decision. Had to refuse. It was a decision not «by logic», but personal choice with doubts, but honest.

Later came another man, who showed himself a little weaker, but there was interest and drive. Now, watching how he works with the team, takes responsibility and fits into the overall structure of the team, I understand: It could not be calculated, only felt.

With all the benefits of AI implementation, the human part of the job should be left to the recruiter. We had a case when the resume of the candidate did not impress the recruiter: poor description, lack of loud companies, design work, technical skills at intermediate level. AI analysis confirmed doubts: the candidate seemed ordinary. But the mentor picked up a man with the words: «There is something in him, we must try». In the interview he spoke quietly and slowly, which at first only confirmed AI’s conclusions. But one detail caught the eye – the candidate clearly and logically described how to restore the system after a failure: without excessive pathos and with a deep understanding of the details. This stood out and at that moment it was intuition that said: This man does not know how to sell himself, but he is reliable. And the team just tired of «stars», which everything is boring for the third month. In the end he passed all stages of interviews and literally in 2 months became a full-fledged combat unit of the team, which gave a very positive feedback.

Ethics, transparency and scalable bias

Artificial intelligence may seem objective. It works by logic, without emotions and personal sympathies, but that doesn’t mean it’s fair. AI is not a moral essence, but a tool that operates on the basis of the algorithms embedded in it. His logic is built on those rules and criteria that the man has set, and if there are prejudices in them, then AI will reproduce them to a greater extent.

We once implemented an AI tool that automatically generates accompanying letters for customers, based on a pre-agreed structure. After a while, the recruiters noticed something strange: letters for men became more confident and decisive, with phrases like «candidate confidently managed architectural tasks», «he has a mature technical eye» and for women softer, as «worked with architecture», «has the knowledge of necessary technologies». The recruiter’s input and resume were identical in level. It turned out that in the training dataset used earlier, strong phrases were used more often in letters for men, and formulations about women were more careful. AI picked up the pattern of behavior from the past and began to replicate it.

What will happen to recruiters in 10 years: spoiler, profession will change a lot

The recruiting profession will not disappear in the next ten years, but it will definitely change a lot. Everything related to mechanical, repetitive actions: resume selection, mass sortation, primary calls and standard interviews with high probability will pass to AI. Leaving specialists who did not seek to develop communication skills, analysis of motivation or strategic thinking.

On the market will remain in demand those specialists who are able to think more widely than a resume and KPI.

The recruiter will become like a psychotherapist with business thinking: a person who can build a deep dialogue between business and the candidate. This role requires other competences: empathy, ability to hear, strategic vision and understanding of team dynamics. It is no longer just the selection of personnel, but the management of human capital. Tools, of course, will change, but The core of the profession – connecting people and opportunities – will remain.

However, it is too early to sound the alarm – despite the buzz around AI in the area of selection, while it performs a supporting role. Yes, technologies are being actively adopted and even small companies are beginning to experiment with automation, but most often these are point solutions built into the individual elements of the system. In fact, AI in recruitment today represents more individual modules within ATS systems than full-fledged through automation. This is comparable to how Excel once appeared in accounting: a useful tool, but no one thought that the profession of accountant had disappeared. Also here – while AI helps, but does not replace.

AI use not only recruiters

According to my observations, there are no more than 10% of specialists on the IT market who completely ignore AI: due to mistrust, fear of «replacement» or ideological rejection. But this is more of an exception than a rule. For the other 90% AI has become a strategic tool, not just a handy clue. This is how it appears when looking for a job:

• Creating resumes and accompanying letters: many applicants do not write them manually. They use AI and adapt the resulting text to the specific job. Some go further and create scripts that automatically replace keywords from the job description. At one of the career consultations, a product manager approached me, asked me to review his resume and showed me a document with a clear, understandable and consistent structure. To my surprise, he said that AI wrote him a resume.
• Interview preparation: neural networks help generate standard questions, formulate «socially expected» answers, train with interviewers and sometimes analyze their own answers.
• Market analysis and strategy: Candidates are asked for AI to compare vacancies, form a list of recurring requirements, assess market salaries or suggest direction of development to remain in demand.
• Portfolio formation: AI helps to write texts for GitHub and LinkedIn, design cases, present projects. It is not so much the search for work, but the creation of such an expertise that the offals come to themselves.

But with good comes also the other side: some specialists use AI as a «cheat sheet» – they take ready-made answers or code, not grasping the meaning. Unfortunately, this practice is only gaining momentum and creating the illusion of expertise that there is no depth. By the way, recently At Habre we talked about IT wolves, who use this mechanism to the maximum. I hope that in the near future there will be tools that will not just «test on AI», but see how a candidate thinks: formulate ideas, solve unusual cases, explains what and why he did. Because now more value is not just the ability to use AI, but the ability to think with it.

AI changes the hiring not only by employers, it also transforms candidates: their behavior, approach to job search and even perception of themselves in the market. Now the applicant who creates a CV with the help of a neural network sends it to a system where at the other end, the same AI is screening. We get into a situation where the machine «talks» to another machine, which changes the rules of the game. In such an environment, the winner is someone who not only knows how to turn on AI, but understands how his logic works from both sides.

Instead of conclusion: how recruiter start to adapt today

To stay in demand, it is important for the recruiter to start rethinking his role. The first step is to develop technological literacy.

At the same time it becomes important to develop that the machine still does not know – soft skills. This is no longer «desirable», but a basic requirement for the specialist in our time. Empathy, the ability to build communication, feel the motivation of the candidate, inspire him – all this comes to the forefront in the competencies of a recruiter.

Separate emphasis on the ability to combine. The recruiter who knows how to use AI as a tool, rather than seeing it as a threat, wins. It is important not to compete with the algorithms, but to learn to cooperate with them, freeing up time for those tasks where a person is really needed. It is the technological flexibility together with emotional competence that will allow the recruiter not only to maintain his role, but to become a professional of a new era.