Środki ochrony praw jednostki wobec wdrażania algorytmicznych systemów decyzyjnych do administracji publicznej – między prawem europejskim a prawem krajowym

The implementation of algorithmic decision-making systems in public administration can pose a risk to the rule of law, democracy, and the right to good administration. Risk management in this regard is intended to be enabled by the instruments provided for in the European draft act on artificial intelligence. The aim of this study is to establish the legal situation of an individual on the grounds of the aforementioned act, and then to relate it to the assumptions of the legislative proposal of the European Law Institute and the provisions of Polish administrative law. The research uses a dogmatic-legal method and a comparative method with regard to references to the common law in force in the United Kingdom. The analysis shows that the European draft regulation completely ignores the individual by not granting him/her legal remedies against the implementation of the said systems. Its protection could be strengthened by allowing complaints to the supervisory authority, whose decisions would be subject to review by administrative courts, as is the case under the provisions of the General Data Protection Regulation. Adoption of these solutions to the Polish legal order could complement the already existing model of control of partly automatic decisions of the administration, providing the possibility to challenge also the principles of their adoption. This would, however, require a determination of who would have a legal interest in lodging a complaint, the possibility for social organisations to participate in these proceedings, and the relationship to other legal instruments already in place, such as public consultations in the legislative process and impact assessments.