Hundreds of AI apps with hidden sexual and violent deepfake features entered Apple's App Store and Google Play Store after passing review. An investigation revealed over 400 apps, seemingly from one Chinese organization, that allow users to generate explicit deepfakes from uploaded photos. These apps masquerade as ordinary services like language tutors or workout coaches.
Under specific conditions, they stealthily load interfaces for creating sexual and violent AI content, including scenarios like "bedroom rape" and "threatening knife." Wired reports the network hides in plain sight via thousands of video ads on Facebook and Instagram. Some apps have existed for a year, passing review by Apple, Google, and Meta despite policies against harmful deepfakes. Developers used bait-and-switch tactics, showing benign content initially but activating hidden features via server checks.
Downloading after clicking Meta ads appears to trigger the explicit content generation. Dozens of pornography templates and credits for deepfake generation are then available. Apple removed 19 apps after Wired's report, stating developers replaced approved content with pornography hosted on third-party platforms.
Apple reviews apps before publication, while Meta has taken action against over 10,000 ad accounts linked to these domains.
Source: 9to5Mac · Summarized by HeadlinesBriefing