Securizarea fluxurilor de date prin platforma MConnect utilizând tehnologii de detecție a amenințărilor bazate pe inteligență artificială

Abstract

This paper explores the integration of Artificial Intelligence (AI) for securing data flows within the Republic of Moldova's Interoperability Platform, MConnect. In the context of the Digital Transformation Strategy 2023–2030 and the rising threats of hybrid warfare, the study analyzes the shift from traditional rule-based security to AI-driven anomaly detection. The research methodology combines a comparative analysis of the Zero Trust Architecture (ZTA) and the technical evolution of the MConnect Events infrastructure. The results highlight that using machine learning models to establish "normative boundaries" for institutional data exchange significantly enhances threat detection accuracy (exceeding 91%) and enables real-time automated response. The study concludes that AI-powered security is essential for achieving cyber resilience and aligning with European Union standards like NIS2 and the AI Act. CZU: 004.056.53:004.8(478); JEL: O33, K24, H11