Meta is yet again under scrutiny as the company ran ads for an AI tool which appeared to promote sexual deepfakes of female US politicians. According to a recent report, users on Meta were displayed ads where one video showed a woman resembling a prominent female politician before cutting to pornographic footage. While Meta platforms have rules banning sexual content and nonconsensual intimate imagery, the ads were reportedly seen across Meta platforms and appeared to be targeted specifically toward male users. The campaign, tied to an app called Kromix, ran for a short period before being pulled. This is not the first time, as Meta has faced repeated criticism in recent months for failing to catch similar ads promoting nonconsensual explicit content, including some involving minors, raising fresh questions about the effectiveness of its automated ad review systems.
The issue was first noticed by Katie Paul, director of the Tech Transparency Project. She says that the ads they found were mainly about sexualising women and using AI to create fake pornographic images of them.
According to the report by Wired, 32 ads linked to the Kromix campaign were found through Meta’s ad library. Kromix was marketed as an AI image styling service, but its advertisements suggested users could create explicit videos featuring real people. The ads ran for periods ranging from five to 46 hours, with most receiving 10 or fewer impressions.
Also read: This country is giving every citizen Rs 30,000 as AI drives its economy
The cover images of the ads often showed harmless scenes such as beaches, mountains and waterfalls, while the videos contained sexual material. The campaign reportedly used the tagline ‘Unleash Next-Gen AI Magic’ and promoted Kromix as an AI tool with ‘no restrictions’. Researchers said the advertisements appeared designed to avoid detection by automated systems.
The app was available to download via Apple’s App Store. However, it was later removed as the Cupertino giant stated that it appeared to have added prohibited features after its initial review. Not only that but the company also added that apps created to generate, distribute or consume pornography are not allowed under its rules.
Also read: Did ChatGPT learn to swear? Users report a sudden rise in F-bombs
The app reportedly allowed users to upload photographs and place people into scenarios including sexual situations. Its content also included AI-generated pornographic videos.
Meta said sexual exploitation and the promotion of nudie apps are prohibited on its platforms. Moreover, it added that it removed more than 340,000 ads promoting nudify services between November and January and is using new AI systems to detect such advertisements.