Mundane AI Apps Hide Sexual Deepfake Tools Behind Harmless Listings
Hundreds of linked apps and domains show how app store review and ad platforms can miss explicit content until it is switched on.
A WIRED investigation found that a group of mundane-looking AI apps, such as a pet portrait generator called Viefur, conceal a network of more than 400 linked domains and apps built to produce sexually abusive deepfakes. The apps appear harmless in app store listings, but when launched under certain conditions they load a hidden web interface that generates explicit AI images and videos from photos users upload. Researchers traced the network to what appear to be Chinese developers and a Hong Kong entity, KeLing Technology Limited, which did not respond to requests for comment. The apps were promoted through thousands of video ads on Meta's Facebook and Instagram, and several were downloaded from Google's Play Store more than 350,000 times collectively. Google, Apple and Meta say their policies prohibit such content, raising questions about how the apps passed review and how quickly enforcement can catch up.
Why it matters
Users who download these apps believing they are harmless can end up with a tool that generates sexually explicit images and videos from their own photos. Because the explicit version is served only under certain conditions, a store listing reviewed at one point may not reflect what the app does later. The findings put pressure on Google, Apple and Meta to show how their review and enforcement catch apps that change behavior after installation, and they leave open how quickly a network spread across hundreds of domains can be shut down.
Matters or hype?
Does this story matter, or is it hype? Vote before you see what everyone else said.