A WIRED investigation published on November 2 uncovered a network of more than 400 connected apps and web domains that disguised tools for generating sexually abusive deepfakes as ordinary services. The deepfake apps posed as pet portrait makers, language tutors and workout planners, and they passed review on Apple’s App Store and Google Play before loading a hidden interface for explicit imagery.
The deepfake apps worked through a server-side switch. After installation, an app contacted a remote server and sent details such as the user’s language, referral source and a device identifier. The server then decided whether to show the advertised feature or load a separate website with explicit deepfake generators. One apparent trigger was downloading an app after clicking an advertisement on Facebook or Instagram.
Researchers found that the deepfake apps’ hidden interface let users upload photographs of real people and place them in sexual scenes, including scenes depicting sexual assault. Templates in the interface were labeled with scenarios such as “bedroom rape” and “threatening knife.” Users could buy digital currency to generate the images, with one app, the pet portrait tool Viefur, selling 130 coins for $4.
How the deepfake apps evaded app store review
The apps passed initial review because the explicit content lived on remote websites, not inside the submitted software. Apple spokesperson Adam Dema said the submitted software complied with company guidelines during initial evaluation, but developers later changed the websites feeding content to the apps, replacing approved material with pornography. Apple said it has suspended tens of thousands of apps over the past year for similar bait-and-switch tactics.
Google removed most of the flagged deepfake apps before WIRED contacted the company, but did not respond to the outlet’s request for comment. In September 2026, researchers identified 27 related apps on Google Play that had accumulated more than 350,000 downloads combined. One app, DiffAI, logged over 100,000 installs before Google removed it.
On October 16, Apple updated its guidelines to strengthen restrictions against nonconsensual intimate imagery and apps that conceal their functionality during review. Guideline 1.2 now addresses AI-generated nonconsensual intimate imagery, requiring developers to prevent or remove such content. Guideline 2.5.19 clarifies the prohibition against hidden, dormant or undocumented features, including functionality activated after approval.
Who researchers traced the deepfake apps to
Researchers at the nonprofit AI Forensics and cybersecurity firm Infoblox traced the network to Chinese-language codebases, Chinese hosting services and a Hong Kong-registered entity called KeLing Technology Limited. Infoblox threat researcher Maël Le Touz mapped nearly 500 linked domains and found the apps shared identical JavaScript configurations and cross-referenced encryption keys.
AI Forensics lead researcher Paul Bouchaud tracked registrations back to late 2025 and found the network registered as many as 30 domain and app combinations per week at its peak. Meta said it had banned KeLing Technology Limited from advertising in 2025 after issuing a cease-and-desist letter. After researchers submitted their dossier, Meta said it acted against 11,600 ad accounts, roughly 4,700 business pages, and blocked 209 apps from its advertising platform.
Apple removed 19 deepfake apps, including Viefur, after WIRED shared its findings. KeLing Technology Limited did not respond to requests for comment, and WIRED reported that the company scrubbed its online presence after being contacted. Meta served thousands of ads promoting the illicit functionality, with some accounts posting marketing materials that linked directly to storefront listings in Apple’s App Store and Google Play.
Why the deepfake apps keep reappearing
The WIRED investigation of the deepfake apps documents removals and enforcement, but WIRED reported that similar promotional campaigns and mirrored applications were still appearing as its reporting ended. The European Union’s ban on nudifier apps takes effect in one month, yet the operators behind the network have repeatedly found ways to push their apps back into stores and ad feeds.
The findings build on earlier research showing that deepfake apps capable of creating nonconsensual sexual imagery were reaching the App Store despite Apple’s restrictions. Previous WIRED reporting found Meta ads depicting AI-generated child sexual abuse material, including images of real children. The new investigation shows the same distribution playbook at larger scale: ordinary-looking apps, remote servers that flip the content after review, and paid ads that steer users to the hidden version.
For users, the warning sign of a deepfake app is one that asks for a photo and then loads a web page instead of working inside the app itself, or an app reached through an ad that promises something its store listing does not mention. Apple and Google both say they have tightened rules targeting abusive generation tools, but the network’s persistence shows how hard it is to keep cloaked software out of the two largest app stores.