The subject of the study consists of procedural limitations of the current model of public-legal protection of citizens in situations where legally significant impacts arise not only as a result of illegal operations with personal data or the adoption of formally expressed administrative decisions but also at an earlier stage – during algorithmic profiling, predictive classification, and the formation of derivative digital characteristics of an individual. The analysis focuses on cases in which computational inference is included in the mechanism of public governance, affects the intensity of administrative control, the volume of managerial attention, or the content of subsequent decisions, yet does not receive an independent procedural formulation. Special attention is given to the limits of the traditional construct of informed consent, changes in administrative discretion when using digital assessments, the difficulties of proving algorithmically conditioned impacts, and the inconsistency of the existing model of liability with the distributed structure of digital causality. The methodological framework comprises formal-legal, systemic, and doctrinal methods applied to analyze legislation on personal data, administrative procedural guarantees, and the legal nature of algorithmically conditioned public influence. A key result of the research is the justification of the category of algorithmically mediated administrative influence. This allows for the qualification of situations in which a digital system does not make a final decision instead of a public entity but, nevertheless, influences its content: it forms a risk assessment, develops a derivative digital profile, establishes a regime for additional verification, or otherwise alters the trajectory of subsequent managerial action. It has been established that the current administrative procedural mechanism primarily recognizes already formalized violations, while a significant portion of the legally significant effect occurs earlier – within the computational procedure and before the emergence of an independent object of dispute. The work distinguishes between the technical opacity of the algorithm and the legal opacity of the impact. The latter manifests not in the complexity of the digital model itself but in the impossibility of reconstructing its role in a specific administrative procedure and verifying how the computational inference influenced the legal status of the citizen. This leads to the necessity for a different procedural logic of protection: a preliminary legal assessment of high-risk systems, subsequent verification of modified algorithmic models, disclosure of the meaning of digital assessment in a specific case, and the possibility of its independent contestation if it has become the factual basis for public influence.
Renat Valer'evich Zanikov (Sun,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: