Name in original language
Otkrivanje poreskih utaja upotrebom veštačke inteligencije
Initiative overview
The initiative focuses on the application of artificial intelligence and big data analytics to improve the detection of tax evasion and strengthen risk-based personal income tax compliance in Serbia. It addresses the challenge faced by tax administrations of limited human resources and growing volumes of complex data by shifting from predominantly random inspections to targeted, evidence-based audit selection.
The project responds to the need for more efficient use of enforcement capacities by analysing large datasets available to the Serbian Tax Administration and identifying atypical taxpayer behaviour that may indicate increased risk of non-compliance. Through the development of algorithms and analytical models, the initiative supports the early detection of deviations from expected patterns, enabling inspectors to focus on cases with the highest risk and potential fiscal impact.
The primary objectives of the initiative are to enhance the effectiveness of tax audits, increase voluntary compliance, and improve revenue collection outcomes while reducing the administrative burden on both the tax authority and compliant taxpayers. By strengthening analytical capabilities, the project aims to support more consistent, transparent, and data-driven decision-making in tax administration, contributing to greater public trust and integrity in the tax system.
The initiative has already demonstrated strong and tangible results, substantially improving audit targeting and efficiency by generating higher newly identified revenues with fewer field inspections. In the longer term, the initiative supports the broader modernisation of the Serbian Tax Administration and aligns with strategic objectives to adopt innovative technologies in public financial management. It also provides a scalable and replicable approach that can be extended to other tax types and compliance areas, reinforcing evidence-based policymaking and the sustainable use of artificial intelligence in the public sector.
