Meta ran ads for an app promising to nudify female politicians
Meta's ad network served promotions for an app explicitly marketing itself as a tool to produce non-consensual nude imagery of female politicians, according to a report from Ars Technica. At least one of the ads included a pornographic video featuring a deepfake image closely resembling an identifiable US politician, meaning the content moved beyond vague promises and into the distribution of explicit synthetic media targeting a named public figure.
The case is a pointed illustration of how generative image technology - now capable of producing convincing likenesses from a single reference photo - can be turned into a tool for targeted harassment. Non-consensual intimate imagery (NCII) involving AI-generated content has become a growing policy and legal challenge globally, with several jurisdictions passing or considering specific laws against it. Politicians, journalists, and other public figures have been disproportionately targeted, and the gendered pattern is consistent: victims are overwhelmingly women.
Meta has long-standing policies against ads that promote adult content, harassment, or the non-consensual use of someone's likeness. The fact that these ads were approved and served - however briefly - points to gaps in automated and human review systems that have struggled to keep pace with the volume and variety of content flowing through large ad platforms. Meta has faced similar criticism in the past over deepfake ads misusing celebrity likenesses to promote financial scams, suggesting this is a systemic weakness rather than an isolated incident.
The broader concern is that advertising infrastructure can act as a distribution and legitimization layer for harmful apps. By paying for placement on a mainstream platform, developers of NCII tools gain reach they would not otherwise have, and the implied association with a reputable ad network can lower a user's guard. Pressure is mounting on platforms to treat AI-generated NCII as a distinct and high-priority content category, with faster detection and stricter pre-approval checks for any app that markets itself around image manipulation of real people.
