News
Poland's Data Protection Chief Calls for AI Rules in Hiring and Firing
Mirosław Wróblewski, head of Poland's data protection authority UODO, has asked the labor ministry to regulate AI systems used to screen job candidates and evaluate employees, warning they can replicate bias and analyze data employers have no right to ask about.
Contents
Mirosław Wróblewski, head of Poland's Personal Data Protection Office (UODO), has asked the Ministry of Family, Labor and Social Policy to begin work on regulations governing the use of artificial intelligence in recruitment and employee evaluation. The move comes amid a growing number of reports about algorithms that automatically screen out candidates and assess employees with little real human oversight.
UODO's request is not a response to a single case but to a broader trend. Recruitment firms and HR departments are increasingly using AI-based tools to scan CVs, profile candidates, conduct preliminary interviews, and support decisions on whether to advance someone to the next recruitment stage or hire them. Similar systems are also being used to evaluate employees already on the payroll.
Risk of replicating bias
The main concern raised by the UODO chief is that AI models are trained on historical data that may itself contain biases related to candidates' gender, age, or background. An algorithm trained on such data can unwittingly replicate and reinforce those biases, producing decisions that appear neutral but are in fact discriminatory.
There is a risk that artificial intelligence systems may replicate biases present in training data. - Mirosław Wróblewski, head of UODO
A second problem concerns the scope of data being processed. According to Wróblewski, AI tools used in recruitment can analyze far more information about a candidate than the Labor Code's defined list of data an employer is entitled to request. Systems that scan social media profiles or content published online may access data that would be unavailable or explicitly prohibited in a traditional recruitment process.
Gaps in Polish law
UODO notes that current data protection law allows exceptions, under national law, to the ban on fully automated decision-making, but requires appropriate safeguards that Polish law does not currently spell out. Among other things, it is unclear who should audit such systems or what appeal process is available to a candidate who believes an algorithm rejected them unfairly.
The office also points to the EU AI Act, which classifies recruitment systems as high-risk. Under Article 86(1) of that regulation, a person subject to a decision made using a high-risk system has the right to a clear and meaningful explanation of the role the system played in that decision.
High-risk systems should be transparent, meaning they should guarantee access to information about how AI models operate. - Mirosław Wróblewski, head of UODO
What it means for employers
For companies already using AI tools in HR, UODO's request signals that the regulatory gap could close faster than some employers expected. If the ministry moves forward with legislative work, businesses could be required to carry out data protection impact assessments before deploying such systems, ensure genuine human oversight of automated decisions, and give candidates a way to appeal an algorithm's verdict.
The AI Act already imposes some of these obligations at the EU level, but UODO wants the details written directly into Polish law, including the Labor Code. Without such clarification, companies may stick to whichever interpretation suits them best, and candidates currently have no clear path to challenge a rejection made by an algorithm.
Broader context of AI-in-HR disputes
The case fits into a wider, international pattern of controversy over AI used to evaluate employees. Similar allegations, concerning algorithmic performance reviews that failed to account for disabilities or maternity leave, have already surfaced against major tech companies abroad. UODO's request does not reference any specific lawsuit, but treats such cases as an argument for regulating the issue in Poland proactively, before the scale of AI adoption in recruitment grows large enough to become a systemic problem.
UODO has offered to provide expert support to the labor ministry in assessing whether future regulations comply with GDPR. It is not yet known whether or when the ministry will begin legislative work, or what shape any new rules would take. The UODO chief's request nonetheless opens a formal path for discussing the limits of automating HR decisions in Poland.

