Cristina ONEȚ
| Abstract: | This article examines the impact of implementing artificial intelligence (AI) technologies within Romania’s tax administration, assessing both the potential benefits and the associated risks. It explores how AI can streamline administrative processes by automating routine tasks, improving data accuracy, and enabling faster analysis of large-scale financial information. These developments have the potential to significantly enhance revenue collection and reduce tax evasion through more effective detection of irregularities, predictive analytics, and real-time monitoring of taxpayer behaviour. At the same time, the article highlights several challenges that must be addressed to ensure responsible and trustworthy use of AI in the fiscal sector. Key concerns include cybersecurity vulnerabilities arising from the increased reliance on digital infrastructures, the protection of personal data in compliance with EU standards, and the need to preserve transparency in automated decision-making systems. Without clear safeguards, the use of AI could undermine public trust or lead to discriminatory outcomes. Drawing on international examples as well as Romania’s specific institutional and legislative context, the study offers practical recommendations for implementing AI in a balanced and legally compliant manner. These insights aim to support policymakers in designing a modern, efficient, and secure tax administration. |
| Keywords: | artificial intelligence, tax administration, digitalization, cybersecurity, transparency |