Former soccer player Ronaldo Nazário announces an online casino game … This video circulates on Facebook. As Engadget remembers, advertising has been seen more than 600,000 times on the platform.
The problem is that it is a deep defake, these hyperrealist videos shared online, which, thanks to artificial intelligence, make anything to say to famous personalities. And the case of the Brazilian celebrity is far from being an exception.
In recent years, more and more stars have seen their diverted image to promote cryptocurrencies or online casino platforms. Conlise Lucet, Lena situations or Emmanuel Macron saw his identity like this Promote cryptocurrency sites or investments.
“Significant amounts of fraudulent content”
However, according to the supervision board, goal (Facebook, Instagram, WhatsApp) is not enough to enforce its moderation rules. Nicknamed “Supreme Court” by Mark Zuckerberg, the Supervision Board, is an agency apart from the company, which does not hesitate to scratch business leaders on moderation issues.
And as indicated, advertising usurping the identity of Ronaldo Nazário has been informed more than 50 times as a scam. However, Facebook, after taking time to eliminate advertising, did not withdraw the associated publication.
Therefore, the Council considers that the company “probably authorizes significant amounts of fraudulent content on its platforms.” He affirms that this matter highlights fundamental failures in the way in which Meta addresses the moderation of the informed fraudulent content that involves celebrities and public figures.
According to the target rules, the company applies to politics “only in case of informing to ensure that personality has not really approved the publication” and that “the interpretation of the concept of false identity” by the moderators could vary according to the regions and cause inconsistencies in the application of the law. “As a result, a large volume of fraudulent content passes between the mesh of the network.
Make facial recognition
Therefore, the Supervision Board urges the group to update its internal guidelines and to give moderators the appropriate media and training to identify deep defects.
For its part, a finishing spokesman says that “many statements from the Supervision Board are simply inaccurate.”
“The scams have gained magnitude and complexity in recent years (…) As this activity has become more sophisticated, our efforts to combat it have intensified. We prove the use of facial recognition, they actively fight against the scams and help users to protect themselves thanks to many safety tools and warnings on the platforms,” adds the spokesman.
In fact, since last March, the company has been testing facial recognition to combat fraudulent ads. Specifically, if the existing moderation systems suspect that an announcement can be a scam, facial recognition comes into play. Then, the device will compare the ads with the photos present in the official profiles of the Instagram and Facebook accounts of the celebrities. If there is a correspondence between the two shots, advertising is automatically eliminated.
Source: BFM TV
