AI-driven digital identity fraud surges in Brazil
Criminal networks in Brazil are using generative AI tools including deepfakes to conduct large-scale digital identity fraud. In the first quarter of 2026, authorities detected nearly 1.5 million identity fraud attempts per Serasa Experian's fraud mapping study, with potential losses reaching nearly R$ 2 billion had they not been blocked.
What happened
In the first quarter of 2026, a new digital fraud risk emerged in Brazil every 13 minutes, an 8.3% increase compared to the same period in 2025, according to the 'Fraude Map' study released by Serasa Experian on 18 June 2026.
Generative AI tools are being used to make scams more convincing, personalized, and difficult to identify. Deepfakes with manipulated audio and video, automated messages, false call center numbers, synthetic profiles, and sophisticated fraudulent pages have become part of criminals' arsenal.
Nearly 1.5 million attempts at identity fraud in cadastral registrations and validations were identified in the first quarter, a 36.6% increase compared to the same period in 2025, representing one attempt every five seconds.
Reported impact
Identity fraud attempts could have caused up to R$ 1.98 billion in losses in the first quarter of 2026 had they not been blocked.
Payment methods experienced the highest volume of identity fraud attempts with 644,500 occurrences, accounting for 43.1% of the total. Telecommunications recorded 313,200 cases, and banks and credit cards recorded 259,100 cases.
Index analysis
Organized networks of fraudsters are now sharing techniques, tools, and vulnerabilities at scale. The number of groups used to share this information grew 139% compared to the first quarter of 2025, reaching nearly 2,000 monitored communities.
More than 19.7 million fraud-related messages circulated among fraudsters in the first quarter, equivalent to 152 messages per minute. Fraudsters operate under a scalable model similar to digital platforms, where criminals identify breaches, package the knowledge, and distribute it to other criminals.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.