Fraudulent ads simulating Brazilian government programs were displayed through the Google platform, even after passing the company's verification process. One example of these materials included the message: 'Congratulations! Your CPF has been approved for Desenrola 2026,' accompanied by an image of a smiling woman and the 'gov.br' symbol, referencing the visual identity of the federal administration, although they were not issued by it.
These scams were detected by researchers from NetLab, a laboratory affiliated with the Federal University of Rio de Janeiro (UFRJ), between August 12 and 25. The fraudsters exploited known public policies, such as Desenrola Brasil, Sistema Valores a Receber (SVR), and CNH Brasil, with the aim of diverting citizens to fake websites.
Analysis and Violations
The findings were detailed in a technical note from NetLab. The researchers clarified that they found no violations of electoral laws related to political party propaganda. However, they pointed out that the ads violate Google's internal guidelines, and the fact that the advertisers obtained platform verification constitutes a governance problem in the advertising infrastructure.
For NetLab, the situation demonstrates that Google uses its advertising infrastructure to benefit advertisers who attempt to impersonate the Brazilian State to deceive users. The researchers located these materials by examining Google's Ad Transparency Center.
Among the elements found, none of the 35 addresses ended in 'gov.br'. However, six of them imitated the design of official government channels, a method known in cybersecurity as phishing. The campaigns also replicated federal symbols and brands, such as Gov.br, Canal Gov, Caixa Econômica Federal, and the country's Coat of Arms.
The messages used were appealing, aiming to attract individuals interested in public programs or solving financial issues. Examples include: 'Your name is on the list to desenroll. You have been selected to desenroll!', 'Value Available in Your Name? See Now,' and 'Free eligibility registration – Free CNH Program.'
The researchers observed that the selection of target programs reflects the scammers' monitoring of official news. Desenrola Brasil, for example, was the most frequent focus; the program, launched in May to help indebted people pay off debts, remained in the fake advertisements until August, when it was closed by the federal government.
Verification and Transparency Functioning
A crucial point highlighted by NetLab is that all identified advertisers had verification granted by Google. This certification is presented by the company as proof of the advertiser's identity and legitimacy. After identity verification and confirmation of the commercial transaction, the payer's name and location are registered in the Transparency Center.
This Center aggregates ads run on services such as Google Search, YouTube, Gmail, and the Display Network, which includes third-party websites and applications that use the company's programmatic infrastructure. In practice, a fake ad can appear in a search and reappear minutes later in a mobile game or a recipe portal.
On Google Search itself, a user typing terms like 'desenrola' or 'forgotten debt' may be exposed to fraudulent content. The researchers also noted that the placement of these ads and the difficulty some users have in differentiating advertising from organic results increase the risk of campaign impact.
Financial Implications and Lack of Transparency
For NetLab, the data suggests that Google profits from fraudulent ads linked to public policies aimed at vulnerable population groups. Desenrola, for instance, is intended for people with financial outstanding balances. Targeted advertising is a pillar of Google's business model; according to the material, the company generated R$ 1.51 trillion just from advertising in 2025, corresponding to over 73% of all revenue collected during that period. However, the criteria used to segment the audience for the ads are not accessible.
Salles commented that, when analyzing how Google promotes its advertising services, the promise is to reach the ideal audience, which is often vulnerable. He explained that part of the logic of micro-segmented advertising consists of finding the user most likely to click, buy, or consume the content.
Another problem pointed out by the researchers is the lack of transparency in the ad query system itself. The Transparency Center does not allow searches by keywords, meaning that fraudulent campaigns can only be discovered if the researcher already knows the name of the advertiser or the promoted website. Furthermore, the tool does not provide data such as the amount invested, the number of people affected, or the option to download the data.
The study also found that 25 of the 50 fake ads related to Desenrola disappeared during the data collection period. There is no record of whether these ads were moderated by Google, removed for other reasons, or simply expired. For NetLab, this lack of records strengthens the hypothesis that the problem is more extensive than what was found in the research.
Without a history of active, excluded, or disappeared ads, it is impossible to identify fabrications that may have misled users while they were being displayed. Although the researchers do not rule out that part of the ads may have disappeared due to Google's own actions against fraud, the scenario remains uncertain.
Regulatory Context
In 2024, the Superior Electoral Court (TSE) implemented new transparency requirements for political-electoral advertising on the internet. In response, Google decided to prohibit this type of advertisement on its platform, exempting it from complying with norms such as creating a specific repository with details on investment, impressions, segmentation, and broadcast period.
Even so, NetLab identified ads from candidates published by Google after this prohibition in 2024. In 2025, the Supreme Federal Court (STF) altered the more flexible interpretation of Article 19 of the Civil Framework for the Internet, establishing a relative presumption of fault for providers in cases of illicit content disseminated via ads or boosting.
In its position, Google stated that it has defined policies to regulate advertising on Google Ads, using a combination of artificial intelligence and human reviewers to monitor ads. The company stated: 'We have clear policies that define the rules for advertising on Google Ads and we use a combination of artificial intelligence, acting preventively, and human reviewers for monitoring. When we identify a violation, we act immediately.'
Additionally, Google reported, based on the Ad Safety Report, that it blocked or removed 374 million ads in Brazil throughout 2025 and suspended 1.3 million accounts during the same period. The company also offered a feature allowing users to report possible violations, as stated: 'We offer a tool for users to contribute and also report possible violations.'