The rise of AI-driven digital fraud demands new identity protection measures

Data from Serasa Experian reveals that nearly 1.5 million fraud attempts were detected in identity registration and verification during the first quarter of 2026—one every five seconds. The report also highlights the misuse of AI, synthetic identities, and fraud-as-a-service as key emerging threats.

The voice as a tool for deception

Criminals can now clone voices using AI tools, generating speech that mimics personal characteristics from brief audio samples. This technology can make fraudulent communications more convincing, as Febraban warned when citing WhatsApp scams that appear to be conversations with known contacts. Additionally, artificial voices are being used to bypass voice-based authentication systems, undermining the assumption that hearing a family member’s or colleague’s voice is sufficient verification.

Visual and data manipulation extends beyond voice

AI can also generate fake images, videos, and other content. In August 2026, the Ministry of Finance reported that a NetLab/UFRJ study identified 544 fraudulent debt renegotiation ads on Meta’s platforms. Of these, 38% were AI-generated, and 72% misused government or brand names and images to appear legitimate. Combining fake identities, manipulated visuals, and urgent messaging in a single approach makes detection even harder.

Institutions must adapt their response to evolving threats

To combat these sophisticated attacks, banks and companies now analyze multiple stages of digital interactions instead of relying on a single verification signal. Febraban stresses the need to strengthen identification methods in response to deepfakes and data manipulation. In 2026, the organization partnered with the Ministry of Management and Innovation and industry associations to expand the use of the National Identity Card in digital verification processes.