Rafał Brzoska and Sebastian Kondracki have unveiled the first results of ScamWatch, a tool that analyses advertising on Meta’s platforms. More than 1.76 million ads were examined, and 16.5 percent of them were labelled „suspicious”. The founder of InPost accuses the tech giant of violating the EU’s Digital Services Act.
The initiative was made public at the end of August and is available at 150proc.pl. Since its launch, the organisers have gathered more than 167,000 signatures under a petition demanding a clampdown on scams across Meta’s services.
Out of the 1.76 million advertisements scrutinised, over 290,000 landed in the „suspicious” bucket, while more than 28,000 were classified as carrying „very high risk”. From the moment the platform went live, its analysts have been reviewing an average of more than 45,000 fraudulent ads per day.
– Today we have more than 167,000 signatures from Poles who want no scams, who want the penalties for breaking the law to be draconian. Our analysis of over 1.76 million ads showed that 16.5 percent of them are highly suspicious ads – said Rafał Brzoska.
The businessman argued that the claim repeated by politicians, namely that Poland lacks implementing regulations in this area, is untrue. In his view, Meta is simply breaching rules that already apply throughout the whole European Union.
– In practice, Meta is violating the provisions of the Digital Services Act, which have been implemented across the EU. This only limits the room for manoeuvre for effective and faster action, but it does not mean a lack of response. Meta is breaking the DSA in several areas – said Rafał Brzoska.
Twelve fraud categories and a 24-hour loophole
The ScamWatch report identifies 12 distinct types of fraud present on Meta’s services and puts forward concrete legislative fixes meant to curb the problem of deceptive advertising online. Its authors also describe the company’s actions as „complicity in a crime”.
According to the document, Meta’s Ad Library is incomplete and operated in a manner that obstructs the fight against criminal advertising. Data is said to appear in the repository with a delay of more than 24 hours, which gives fraudsters room to swap a fraudulent ad for an ordinary one and conceal it from Meta’s own systems. The platform is also said to permit ads to be modified after publication.
– This is a deliberate opening of the door for criminals so they can publish their fraudulent ads and then disappear from the virtual space. Meta lets them replace an ad – for example a deepfake with a jeans advert – after publication. The criminal carries out the procedure because Meta allows it. Then he swaps the ad so that the trace of the procedure disappears from the library. We managed to capture and prove this – said Rafał Brzoska.
Another argument pointing to breaches of the DSA concerns independent organisations, which the report says are hindered because they cannot download advertising data in bulk. The Instrat Foundation had already reported that only 50,000 records from a single day can be retrieved, while fraudulent ads may run into the millions.
Meta denies the findings, Brzoska invites a lawsuit
Instrat calculated that the corporation could be earning as much as 760 million złoty a year from fraudulent advertising. Meta responded by calling that report „fiction created to order, which serves to mislead public opinion”, adding that it had been commissioned by a law firm representing a person who sued the company.
Brzoska rejected those arguments as absurd and said that if Instrat’s numbers were false, Meta would publish its own calculations. He also claimed that ads featuring him practically vanished only after the 150 proc initiative was launched a few weeks ago, and not following court rulings.
– They will not do it. It is a sham that they fight scams and present themselves as „fighters removing scams”. Ads with me practically disappeared, but only after the 150 proc initiative was launched a few weeks ago, not after court verdicts – said Rafał Brzoska.
– If they do not agree with this, I encourage them: sue me. They will not do it, because they would have to disclose documents the court would demand: about the real wave of scams and show how they really fight fraudulent ads. That is why they stay silent. Hence a scam ad is 3.7 times more expensive than a standard ad, because they agree to it. It pays off for them – said Rafał Brzoska.
Sebastian Kondracki of Bielik AI admitted that he has to watch suspicious advertisements personally before reporting them manually for removal in Meta’s systems, since there is no option to report fraudulent ads on a mass scale. He also pointed to a separate stream of very low-quality content appearing on the platforms.
– I have the unpleasantness that I have to watch any suspicious ads myself, sometimes watch the whole application. I will stress that such a lack of safeguards makes Meta a huge factory of ads – strongly suspicious or with elements of scam. However, a separate group is also made up of very low-quality ads: Chinese AI novels, AI porn materials presenting – I will stress – either very young women or even children. This is about applications for generating sexual deepfakes, which contain instructions how to do it, and which are allowed from the age of nine – said Sebastian Kondracki.
The authors of the analysis intend to send their report to every MP, senator and member of the government, so that, as Brzoska put it, they know what they are consenting to. Kondracki added that work is already under way to extend the fraud monitoring to X and TikTok as well.
Źródło: WNP.pl, Fot. Shutter Speed/Unsplash; 150proc.pl
