More than 400 AI apps and linked domains were found hiding abusive deepfake tools. The apps were available on Apple’s App Store and Google Play Store. Developers presented them as ordinary utilities, while hidden web interfaces created explicit content and threatening videos from uploaded photos, a WIRED investigation published on November 2, 2026, found.
The network of connected apps showed harmless features to platform reviewers at first. Some promoted pet photo editing, language lessons, and AI fitness coaching. Specific triggers later opened separate interfaces with tools to generate non-consensual sexual images and violent deepfakes.
The developers used a bait-and-switch approach to pass safety checks. The visible features looked genuine during review, while remote servers controlled access to the hidden tools. Ads on Facebook and Instagram also pushed these apps toward potential users.
Some hidden tools offered templates showing sexual assault and physical threats. Users could upload a photo, pick a scenario, and buy credits to generate manipulated images or videos. Such features raise serious concerns about harassment, humiliation, blackmail, and threats against targeted people.
Apple responded after WIRED shared its findings. The company removed 19 identified apps from the App Store, including Viefur. Apple said it reviews every app before publishing, and these apps followed its rules at that time. According to the company, developers later swapped approved content for pornographic material and promoted it through third-party ads.
Google also acted against apps that broke its policies. Meta reported action against more than 10,000 advertising accounts tied to domains linked with the network.
The network had a wide reach across app stores and ad platforms. By September 2026, 27 related apps had gathered over 350,000 downloads on Google Play. One app, DiffAI, reportedly passed 100,000 downloads before its removal.
Meta later shared wider enforcement details. The company acted against 11,600 advertising accounts, 4,700 business pages, and 15,000 pages linked to the apps. It also blocked 209 apps from advertising on its platforms.
These findings point to a growing challenge for digital platforms as AI image tools become easier to distribute. App reviews must now consider remotely controlled features, not only the functions visible at submission.
Also Read: Tamil Nadu Police Alert: Vijay Deepfake Scam Goes Viral