Brazil bans paid ads for deepfakes in final 72 hours…
Brazil's data protection authority published a technical report detailing how deepfake technology is being used to conduct large-scale financial fraud and…
- Incident date
- Jul 2026
- Target
- Gisele Bundchen, Angelica, Juliette, Sabrina Sato, Drauzio Varella, Pedro Moreira Salles, Brad Pitt, Sergio Moro
On July 29, 2026, Brazil’s national data protection authority (ANPD) released a 163-page technical report mapping the rising prevalence of deepfakes in financial fraud, electoral manipulation, and non-consensual imagery. While the document carries no binding legal force, it provides a comprehensive overview of how synthetic media is being weaponized to target consumers and influence public discourse.
What happened
The report identifies a recurring three-step operational pattern used by criminal groups: the creation of synthetic content featuring the voice and image of public figures, the boosting of this content via sponsored social media advertisements, and the redirection of victims to fraudulent websites to extract payments or personal data.
One major case cited involved celebrities including Gisele Bundchen, Angelica, Juliette, and Sabrina Sato. Criminals used deepfakes of these figures to promote non-existent products, such as a fake anti-wrinkle kit. Victims were charged shipping fees via Brazil’s PIX instant-payment system, resulting in more than R$20 million in illicit gains before the group was arrested. Similarly, the likeness of physician Drauzio Varella was repeatedly manipulated to simulate medical endorsements, often mimicking news segments to gain credibility.
Beyond advertising fraud, the report documents instances of financial impersonation. One case involved a video using the image and speech of Banco Itau co-chairman Pedro Moreira Salles to promote a fraudulent investment scheme. In another instance, a woman in Rio Grande do Sul engaged in real-time video calls with a synthetic version of actor Brad Pitt. The report also highlights the exploitation of the Serasa Experian brand, where deepfake videos of former senator Sergio Moro were used to lure victims into believing they were eligible for compensation related to a 2021 data leak.
These findings align with broader industry data, such as a 126 percent increase in deepfake attacks between 2024 and 2025. Platforms including Meta, Google, and Microsoft have attempted to counter these threats through mass account suspensions and improved detection algorithms, though the report emphasizes that the harm is defined by the scale of many small losses, which often discourages victims from filing formal complaints.