Faculty of Social Sciences, University of the National Education Com-mission in Krakow
ul. Podchorążych 2, 30-084 Kraków
e-mail: amadera@poczta.onet.pl
THE RISK ASSESSMENT PROCESS RELATED TO THE USE OF ARTIFICIAL INTELLIGENCE SYSTEMS IN LIGHT OF THE ARTIFICIAL INTELLIGENCE ACT
Abstract. This article analyses the process of risk assessment related to the use of artificial intelligence systems under the European Union Artificial Intelligence Act. The main purpose of the article is to explain how the classification of an AI system determines the scope of legal obligations imposed on providers and deployers. The article argues that the AI Act does not establish a uniform regulatory model for all AI systems, but in-stead adopts a graduated, risk-based approach. Under this model, the legal consequences depend primarily on the system’s intended purpose, speci-fic use case, context of application, and potential impact on natural per-sons. The article concludes that risk assessment under the AI Act is not merely a technical or formal compliance exercise, but a functional and contextual legal process. Proper classification requires an active asses-sment by the entity intending to use the system, including an analysis of its purpose, users, data, decision-making role, and possible effects. Only such an approach allows organisations to determine whether an AI system is prohibited, high-risk, limited-risk, or minimal-risk, and to ensure its lawful, responsible, and safe use
Keywords : AI Act; artificial intelligence; risk assessment; high-risk AI systems; transparency obligations.
JEL classification: D80,D81