Is There a Legal Obligation to Use Artificial Intelligence in the European Union Accession Process?

This research aims to examine whether there are any legal grounds to assert a responsibility on Serbia’s part if the country does not utilize artificial intelligence (AI) in the process of joining the European Union (EU) even when there is no textual reference to such obligation. To comprehensively explore potential legal grounds for invoking this particular duty, we must first establish the applicable legal frameworks in Serbia. We shall investigate the formal grounds of Serbia – European Union association process, which were established by the Stabilisation and Association Agreement, in the first part of the study. The focal point of the inquiry revolves around the significance attributed to this agreement within the Serbian domestic legislation. The objective is to determine if Serbia can potentially avoid its international obligation to employ AI based on its internal laws. This will help us in identifying the nature of the obligations Serbia may possibly have in this context, as well as the general principles guiding their interpretation. Subsequently, we will analyse the specific obligations that Serbia, as a candidate country, currently possesses during the integration stage. In the next section of the paper, we will highlight the inadequate fulfilment of some of the main requirements of the process, in order to better understand whether AI could assist in meeting these obligations and what characteristics such AI would have. This will lead us to the central research question at hand – whether there is a duty to utilize AI. We examined the rules pertaining to the possibility of evasion due to impediments, in contrast to the principle of good faith performance, in order to draw our conclusions.