Scammers use government programs in false advertisements on Meta: Valors a Receivable and Minha Casa Minha Vida are on the list, points out NetLab/UFRJ

For Débora Salles, general research coordinator at NetLab, the number represents just one part of a larger, organized structure. "We call it the disinformation industry, because we see that it is an organized and structured business. Although 860 ads is already a considerable volume, we know that we are seeing the tip of the iceberg. We face many limitations in the transparency that Meta offers about advertising. So, what we were able to collect and archive is a fraction of the problem", he explains. "A single ad can be seen by thousands or even millions of people. When we count an ad, this does not mean that it is just a post seen by a few people. We do not know how many people saw this content, because Meta does not offer this type of information either", says Débora. When contacted by g1, Meta did not respond until the publication of this report. The service most used by scammers was Valors Receivable, from the Central Bank, mentioned in 345 advertisements. Then they appear: Brasil Smiling: 230 ads; My Home My Life: 67; Gas Aid: 62; Bolsa Família: 31; Brazilian driving license: 30; Work Card: 29; Desenrola Brasil: 19. According to the survey, 26.3% of advertisers explored more than one public policy in their publications. According to Débora, programs that involve direct payments or benefits have greater potential to attract victims. Valors a Receivable also reaches a wide audience, without restrictions on age or socioeconomic status. "When it involves receiving money, there is a question of being more attractive to the user: 'Oops, I can get help easily, help that I didn't even know I was entitled to'. This works as a lure for the possible victim. In the case of public policies, the more direct the financial relationship, the greater the appeal." The pieces used false promises or out-of-context information to make users believe they were entitled to non-existent money, benefits, or services. In some cases, the ads directed victims to websites that charged a fee to consult the supposed benefit. One of the false promises stated that the government would have released up to R$12,000 for all people who worked in recent years. Other ads advertised a non-existent cashback program for credit card users. The ads also appropriated the identity of press vehicles and content with a journalistic appearance to appear legitimate. Among the cases identified, fake pages used the name and visual elements of g1 to publicize supposed queries for available values. The parts were unrelated to the vehicle and sought to take the user to fraudulent websites. Websites simulated official channels The study identified 185 advertisements, 21.5% of the total, that used pages and websites to simulate official government channels. In 154 ads, the user was directed to a website that presented itself as an official page. Another 66 pieces were published by pages that used names or visual elements associated with the government and public programs. The two strategies appeared combined in part of the publications. Website addresses also included references to public institutions and services, as well as expressions such as “official” and “org”, to convey legitimacy. In one of the examples analyzed, a page promised a free consultation with Brasil Sorridente, but sent the user to a website full of advertisements. According to NetLab, the objective was to make people scroll through different pages to generate advertising revenue. According to NetLab, many of the pages practically do not maintain regular public activity and are reused for different campaigns. "Often, they have profiles that are almost ghosts on the platforms. From the point of view of organic content, they publish very little. They also change their name and image: at some point the page was Fabiana, then it became gov.br and then it became Ronaldo. They are recycled", he explains. Almost half used artificial intelligence Of the 860 ads analyzed, 394, or 45.8%, featured videos generated or manipulated by artificial intelligence. Among the strategies identified are videos with manipulated images of politicians, journalists and other public figures. According to the research, these familiar faces were used to give credibility to publications and mislead users. A deepfake of federal deputy Nikolas Ferreira (PL-MG) appeared in at least 16 ads in the sample. According to NetLab, the video had been used since 2025 in scams related to Pix. Fraudulent advertisements used a deepfake of congressman Nikolas Ferreira and a fake page with the name and visual identity of g1 to publicize supposed amounts to be received Reproduction/NetLab-UFRJ; Arte/g1 The study also found 146 ads, 17% of the total, that used Meta's own “dynamic creative” feature. The tool combines different images, titles and texts with artificial intelligence to show different versions of the same advertisement. Thus, an ad can display different content depending on the user reached or the platform.