News
Poland's Hiring Process Turns Into a Duel of AI Algorithms
A Pracuj.pl report finds that nearly half of job candidates in Poland use AI during recruitment, while only 7 percent of companies have clear rules governing its use. The result is a hiring process where a candidate's algorithm increasingly competes against a recruiter's algorithm, blurring the line of fair play.
Contents
Poland's labor market is entering a phase where hiring decisions increasingly hinge on a clash between two algorithms rather than a conversation between two people. A report from Pracuj.pl shows that nearly half of candidates now lean on artificial intelligence at some stage of the recruitment process, while the vast majority of companies have no written rules governing its use.
The phenomenon was described on Wednesday by the outlet iMagazine in a piece titled "Candidate versus recruiter, or how AI started recruiting AI," citing data from the Pracuj.pl report along with comments from labor market experts. The authors point to a paradox: artificial intelligence was supposed to make candidate screening easier, but in practice one algorithm increasingly clashes with another before any human gets the chance to assess actual competence.
An algorithmic arms race
On the candidate side, AI is used today mainly to tailor CVs and cover letters to a specific job posting, to prepare for interviews, and to research prospective employers. 43 percent of respondents see the use of such tools as resourcefulness rather than cheating.
On the other side, recruiters are increasingly relying on systems for initial application screening, automatic candidate ranking, and analysis of answers submitted through forms. The scale of this process means that hundreds of applications produced by similar language models start to look nearly identical, making it harder for recruiters to distinguish real competence from a skillfully written text.
Where the line of acceptance sits
The Pracuj.pl report shows that recruiters' acceptance of AI varies sharply depending on the stage of the process. 67 percent of recruiters tolerate the use of AI to prepare for an interview, and 57 percent accept its help in drafting a CV. Approval drops sharply when it comes to the recruitment task itself, with only 19 percent of respondents viewing this positively, and just 3 percent accepting answers generated live during the interview.
Companies should build clear AI policies for candidates. For example: research, preparation, and drafting a CV are allowed. What's not allowed are answers generated by AI during the interview itself - Kacper Lasota, Head of Talent Acquisition, XTB
Lasota also notes that the real problem isn't the technology itself but the methods of evaluating candidates, which haven't kept pace with its development. In technical recruitment, recruiters quickly spot when an engineer with no knowledge of a given technology tries to read an answer off a screen, which in his view calls for rebuilding how competencies are verified rather than fighting AI itself.
Employer brand inside language models
The report highlights another neglected area: how companies are portrayed by the language models themselves. Only 2 percent of recruiters regularly check how large language models describe their employer, and 61 percent of companies take no action on this front at all. Since candidates increasingly ask chatbots for opinions about a prospective employer before applying, having no control over what AI says about them is becoming a genuine reputational risk.
What the AI Act changes
The regulatory context adds further pressure. The EU's AI Act classifies systems used for recruitment and candidate selection as high-risk solutions, which means decisions made or supported by algorithms must be subject to proper human oversight. The report's authors stress that at many companies this oversight remains a formality today, while real decisions increasingly get made at the automatic application-screening stage.
The lack of written rules on both sides of the recruitment process leaves candidates and employers operating under uncertainty. A candidate doesn't know whether admitting to using AI will hurt their chances, and a recruiter often has no way to even determine that they're dealing with machine-generated text or answers.
Pracuj.pl's experts stress that the labor market has reached a transitional moment in which the growing role of technology is complicating recruitment processes rather than streamlining them. In their view, clear and communicated rules on AI use should form the basis for building transparency and mutual understanding between candidates and employers.
For Polish companies, this translates into concrete work to complete before the end of the year: writing an AI policy for recruitment processes, clearly defining at which stages algorithmic assistance is permitted, and training recruiters to recognize machine-generated content. Without this, they risk a situation where hiring is decided by the incidental effectiveness of one algorithm against another rather than by a candidate's real competence.
