Artificial Intelligence in Public Authority Decisions: Constitutional and Legal Limits of Algorithmization

The subject of the study is the constitutional and legal limits on the use of artificial intelligence and other algorithmic systems in making public authority decisions that affect the rights, freedoms, and legitimate interests of citizens. Particular attention is paid not to the technical aspects of algorithm functionality, but to the legal consequences that arise when the result of automated data processing effectively influences the provision, restriction, or termination of rights, the assignment of liability measures, denial of social benefits, access to public services, or any other change in an individual's legal status. The research focuses on the problem of maintaining legality, reasonableness, individual approach, and judicial protection in situations where the actual logic of a decision may be formed within an opaque digital system. The subject also includes an analysis of procedural guarantees necessary to ensure that the application of artificial intelligence in public governance does not replace legal assessment with automated conclusions. The methodological framework of the study comprises formal legal, comparative legal, systemic, functional, and legal modeling methods; additionally, it employs the analysis of foreign case law and international acts in the field of artificial intelligence. The scientific novelty of the research lies in the proposal to examine the problem of artificial intelligence in public governance not through the traditional question of the legal personality of artificial intelligence, but through the category of constitutionally significant algorithmic decisions. This approach allows shifting the focus from abstract discussions about the status of technology to the real risks for individuals arising from the use of algorithms by public authorities. A new aspect is also the suggestion to link the level of legal guarantees not to the name of the technology but to its function and consequences for the citizen. The research concludes that the use of artificial intelligence in public governance does not in itself contradict constitutional principles; however, it requires clear legal limits and specific procedural guarantees. The main danger of algorithmization lies not in the autonomy of technology, but in the potential for a citizen to face an unfavorable public authority decision, the actual logic of which is hidden in data, models, or automated classifications.