PCDF Busts Deepfake Donation Fraud in Brazil

The Federal District Civil Police has broken up a criminal organization suspected of raising nearly R$10 million through fake donation campaigns online. According to the force, the group used artificial intelligence, including deepfakes, to produce fake videos of Gusttavo Lima, Marcelo Rossi and Fábio de Melo, then redirected users to pages that imitated legitimate charity drives.

How did the scheme work?

Police investigations indicate the suspects used AI-manipulated videos to push people into making financial transfers. The materials featured public figures, but also images of vulnerable people, with the goal of building empathy and speeding up donations, according to O Globo and G1.

G1 also reported that the group used artificial intelligence to create videos of fake elderly people, children and people with serious illnesses. The tactic was meant to move victims emotionally and lead them to donate to pages that imitated real solidarity campaigns.

What amounts did the investigation attribute?

The PCDF attributed nearly R$10 million to the group, an estimate tied to the investigation and not a judicially confirmed figure. Diário do Centro do Mundo reported, based on the police probe, that a company linked to the scheme received 215,575 transactions totaling R$9,839,087.73 in 45 days.

Metrópoles also reported that the PCDF identified 66 police incident records in the Federal District tied to companies flagged as recipients of funds linked to electronic fraud. That figure broadens the geographic scope police observed in the case.

What other cases show AI use in Brazil?

That same day, the Ministry of Science, Technology and Innovation released a separate case of AI use in cybersecurity. The Hackers do Bem program presented a system developed by the GT-MALWARE DATALAB at the Federal University of Pampa to generate synthetic data for training malware detection tools.

The initiative also combines data anonymization, classification through language models and automatic generation of playbooks and training paths. In this case, AI is being used to support defense, while in the PCDF investigation it appears as a tool for scaling fraud.