Séries históricas Itaú Unibanco warns about deepfake scams and reinforces digital security guidance
As the use of Artificial Intelligence in fraud attempts grows, the bank highlights protective measures and reinforces precautions to help clients avoid scams
São Paulo, June 2026 – The advance of Artificial Intelligence has increased the sophistication of digital fraud attempts in Brazil and around the world. Among the scams gaining prominence are those that use deepfakes, voice cloning, and other forms of image, video, and audio manipulation to simulate real situations and induce victims to make transfers or share personal data.
Attentive to this scenario, Itaú Unibanco is strengthening its work in preventing scams and fraud, combining technology, behavioral analysis, data governance, and ongoing awareness efforts to support clients and society in confronting this type of threat.
According to the global FBI IC3 Crime Report, losses associated with digital crimes reached US$ 20.8 billion in 2025, the equivalent of about R$ 115 billion. Of this total, 85% were related to digital fraud, and nearly US$ 893 million, approximately R$ 4.9 billion, involved some type of Artificial Intelligence as part of the criminal approach.
“Artificial Intelligence tools have made scams more sophisticated, especially when combined with social engineering techniques. Often, the criminal’s goal is not to create a technically perfect fraud, but to generate urgency, trust, or fear to lead the victim into making a hasty decision. That is why we act preventively, with layers of security that analyze context, behavior, and risk signals before and during the client’s journey,” says Felipe Tambelini, Fraud Prevention director at Itaú Unibanco.
In Brazil, attacks involving deepfakes grew 126% in 2025, while impersonation scams, in which the criminal poses as another person, rose 140%, according to data released by Biocatch. The topic also concerns consumers: 89% of Brazilians say they fear having their voice cloned by Artificial Intelligence tools.
“We continuously invest in technology and intelligence applied to preventing scams and fraud, always considering the reality of our clients and the main types of scams observed in the Brazilian market. In addition to security solutions, we are also committed to guiding the public, since information is one of the main forms of protection,” adds Tambelini.
How AI scams work
Criminals have been using Artificial Intelligence tools to create fake audio, video, images, or