This report presents the legal considerations that one should take when using data and algorithms in public employment services. It largely draws on examples from statistical profiling as this is currently one of the most common algorithmic tools used in public employment services (PES). Importantly, the general lessons are applicable beyond profiling for example for the purpose of data-driven decision support systems in PES. The report discusses the legal basis for algorithmic decision making and among other discusses proportionality, discrimination, fairness and protection of sensitive data. It highlights the necessity to conduct a thorough assessment of issues such as data quality, accuracy and transparency during the whole life cycle of an algorithmic tool. The report also intends to provide lessons for the HECAT algorithmic pilot tool
Næsborg-Andersen et al. (Fri,) studied this question.