Unico warns AI-enabled sophisticated fraud attacks could spike 550% in 2026

Most observers of digital fraud would grant the likelihood that sophisticated fraud attacks will increase this year, the question is: by how much? The answer is somewhere between 150 percent and a whopping 550 percent , according to research from Unico.
These insights were shared by Unico Technical Adviser Davi Reis during a presentation on “Authenticity vs. Deepfake: Who Validates What’s Real?” at SXSW 2026.
The company observed a 1,082 percent increase in digital fraud attempts in 2025, compared to the year before. And sophisticated fraud is becoming faster and cheaper to attempt, Reis says, with the wide availability of AI tools. These have lowered the barrier to entry, with Unico finding that attacks that previously required 5,000 Brazilian reals (roughly US$955) worth of hardware now possible with software subscriptions costing less than R$30 ($5.70) a month.
The company drew its data from more than 1.5 billion authentications performed by its biometric identity verification customer base, which is mostly concentrated in Latin America. The company claims a 0 percent False Acceptance Rate (FAR) in those authentications, indicating all fraud attempts were blocked.
“As deepfakes become more realistic and accessible, protection relies less on visual analysis of the content and more on understanding the context,” says Reis. “From a structural standpoint, technologies already exist — such as those from Unico — that can mitigate this risk, especially in regulated environments like banks, large retailers, and digital platforms, which adopt verification solutions at the moment the video is captured, before any fake content can be used. The trend is for these technologies to become increasingly common, expanding consumer protection as deepfakes evolve.”
Unico recently published a white paper by the same title as Reis’ presentation in collaboration with Biometric Update, which is available for free download with registration.
Article Topics
AI fraud | biometrics | deepfakes | digital identity | fraud prevention | generative AI | Unico






Comments