AI-driven advanced digital fraud surges 180% in Romania in 2025
Advanced digital fraud cases increased 180% year-over-year in Romania in 2025, driven by widespread use of AI tools to create synthetic identities, deepfakes and autonomous attack systems. Sophisticated fraud now accounts for 28% of all fraud cases, up from 10% in 2024, with phishing and supplier breaches remaining common attack vectors.
What happened
Advanced digital fraud cases grew 180% in 2025 compared to the same period in 2024, according to data from identity verification company Sumsub cited by Romania's National Directorate of Cybersecurity (DNSC).
The share of sophisticated fraud increased from 10% to 28% in a single year. Attackers increasingly use autonomous systems capable of executing multi-step fraud with minimal human intervention, and AI-generated synthetic identities. Phishing remains the most frequent attack method at 45%, while 36% of fraud comes from breaches at suppliers.
AI-generated documents represented 2% of fake identities and records used in digital fraud in 2024, powered by tools including ChatGPT, Grok and Gemini, representing a concerning upward trajectory. In contrast, the USA recorded a 15% decline in overall fraud rates year-over-year, though 21% of incidents there involved synthetic identities or AI-generated personas.
Reported impact
Cybercriminals have shifted from high-volume opportunistic attacks to AI-driven precision operations that are harder to detect and cause substantially larger damage.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.