ПОТЕНЦИАЛ ЗА ИЗПОЛЗВАНЕ НА ИЗКУСТВЕН ИНТЕЛЕКТ ЗА РАННО ОТКРИВАНЕ НА КОРУПЦИОНЕН РИСК ПРИ ОСЪЩЕСТВЯВАНЕ НА ДЕЙНОСТ В ПРИХОДНАТА АДМИНИСТРАЦИЯ
POTENTIAL FOR THE USE OF ARTIFICIAL INTELLIGENCE FOR EARLY DETECTION OF CORRUPTION RISK IN THE PERFORMANCE OF ACTIVITIES WITHIN THE REVENUE ADMINISTRATION

Summary/Abstract: This scientific publication examines a current and under-researched issue of high practical relevance, focusing on how public administration - and particularly the revenue authorities - can implement algorithms and models of artificial intelligence for the early detection of corruption risk. The study emphasizes the operational model of application - from the selection of data and indicators, through the integration of algorithms into administrative processes, to the boundaries of accountability, human oversight, and evidentiary traceability of the actions undertaken. A practical framework for the implementation of artificial intelligence is proposed, grounded in the principle of “integrity by design”. This framework combines the requirements of European Union law and the national Bulgarian legal framework with applicable procedural solutions in the day-to-day work of the revenue administration. From a practical perspective, the paper supports the thesis that the most sustainable approach is to use artificial intelligence as a tool for prioritization and secondary control within established administrative procedures, rather than as a standalone decision-making mechanism that could lead to negative consequences and an adverse outcome. The expected effect of developing and introducing a model such as the one proposed is twofold: on the one hand, increasing the level of early detection of corruption indicators (procedural deviations, unusual exceptions, recurring dependencies), and on the other hand, enhancing institutional resilience through transparent, verifiable, and standardized rules of action compatible with European and national regulatory requirements.

• Details
• Contents