In this study, we address the circulation of fraudulent ads which use images of Brazilian religious leaders. These ads frequently contaied audio and video manipulated by artificial intelligence (AI).
The data was collected with Meta's Ad Library interface and includes ads published between July 1 and 20, 2024.
The ads promote scams by displaying the image of pastors, such as Claudio Duarte and Silas Malafaia, and priests, such as Fábio de Melo, Marcelo Rossi and Reginaldo Manzotti.
Most of the ads promote products with no proven efficacy aimed at various health issues, such as pain, sexual impotence and weight loss.
While they appropriate images and exploit the popularity of priests and pastors with AI-generated content, as of the time of writing this report, no ads had been moderated by Meta or labeled to signal the use of AI.
The analysis points to the increasingly widespread application of AI to manipulate information and create misleading or fraudulent content.
In addition, it suggests that Meta’s advertising service may not restrict the use of sensitive personal data – such as information about religious beliefs – in ad targeting, which would constitute a serious systemic risk arising from the company’s operational strategies.
Click below to download the document
Note: The report was edited on 10/09/2024. Some parts were removed to accommodate the information in the current version.
DATA ACCESS
If you wish to access the database for this report, please send an email to netlab@eco.ufrj.br identifying yourself, explaining why you are interested in the data and how it will be used. NetLab will evaluate your request and contact you.
WARNING
This report is an independent production of NetLab UFRJ. All decisions regarding this work were made exclusively by the researchers of the laboratory. The funders of NetLab UFRJ have no influence over the laboratory's research agenda and did not participate in any stage of the production of this report.
Information about NetLab UFRJ's funding sources is available here.