top of page

AI ads use images of Brazilian politicians to carry out scams



At the end of January, scams using Artificial Intelligence (AI) involving a supposed compensation pay out by the company Serasa began to circulate in advertisements on Meta's platforms. According to UOL, the ads gained traction from mid-February.


Despite reports in articles from the main press vehicles and checks carried out by Reuters and Aos Fatos, as well as Serasa itself denying the payment in an official statement, the scam continues to be broadcast by Meta.


With a view to understand how the scam has persisted and the use of AI, we collected advertisements for Serasa from the Meta's library. 


From the content analysis, we identified 4,321 advertisements that ran the scam and that used AI tools to generate the content.


What is not included in the articles and checks, and which we identified in our report, is the use of public figures from different parties, on both sides of the political spectrum, to scam consumers. All identified advertisements run scams using excerpts from videos of deputies Nikolas Ferreira, Sergeant Fahur, Eduardo Bolsonaro and President Lula, manipulated using AI to deceive consumers. 


We also identified numerous pages and profiles with signs of inauthentic behavior and sites created to simulate official channels, such as that of the Federal Public Ministry.


Read the report


Anúncios com IA usam imagem de políticos brasileiros para aplicar golpes
.pdf
Download PDF • 5.78MB

Institutional
Contact
logo_atualizada_branca.png
assinatura.png
ufrj-horizontal-negativa-completa-telas.png

© NetLab UFRJ 2023.  This work may be freely copied for non-commercial teaching and research purposes. If you want to make any other uses that infringe copyright, contact our coordination by email.

bottom of page