Meta allowed obscene advertisements that break its content moderation rules

In 2024, Meta allowed over 3300 pornographic advertisements-AI-related material-based content-on their social media platforms, including Facebook and Instagram.

Conclusions come from one Report By AI forensicA European non-profit organization focused on investigating the technical platform algorithm. Researchers discovered an incompatibility in the meta content moderation policies by re -uploading several clear images as standard posts on Instagram and Facebook. Unlike advertisements, those positions were rapidly removed for a violation of meta Community Standards,

“I am both disappointed and not surprised by the report, given that my research has already exposed the dual standards in content moderation, especially in places of sexual materials,” Carolina is At the center of the University of Northerbly for digital citizens in Britain.

The AI ​​forensic report focused on a small sample of advertisements aimed at the European Union. It was found that clear advertisements allowed by Meta mainly with a total access of more than 8.2 million raids, with a total access of more than 8.2 million raids, with a total access to “suspected sexual growth products” and “hook-up dating websites” to old men and old men. .

Such permission refers to a wide dual standard in the material moderation. Tech platforms often block materials for “women, fame-pre-presenting and LGBTQA+ users”, she says. This double standard male and female sexual health extends. “An example is advertising from lingerie and duration from the meta (removed), while advertisements for Viagra are approved,” she says.

In addition to finding the AI-borne imagery in advertisements, the AI ​​forensic team also discovered audio deepfec: in some advertisements for sexual growth drug, for example, pornographic scenes were overladed with a actor Vincent Castle’s digital manipulation voice.

“Meta prohibits the performance of nudity or sexual activity in advertisements or organic positions on our platforms, and we are removing violations shared with us,” says a Meta spokesman says. “Bad actor is constantly developing his strategy to avoid enforcement, which is why we continue to invest in best equipment and technology to help identify and remove violated materials.”

Report matches the CEO of Meta Mark Zuckerberg that the company will get rid of its fact-oriented teams in favor of crowded community notes.

“If we really want to sound the diastopian-and at this stage Zuckerberg’s latest decision has been given to remove facts, then I think we can also say that we have the reason for being. That meta is a hurry to take money from its agency’s personal, marginalized users.