How ordinary apps could conceal abusive deepfake tools

WIRED reported on November 2, 2026, that more than 400 connected apps and domains presented as ordinary services while potentially exposing hidden tools for making abusive AI-generated sexual images and videos. The reported mechanism relied on a server choosing which version of an app to show.

How an ordinary-looking app could reveal a different interface

AI Forensics researcher Paul Bouchaud told WIRED that an app contacted a server with information such as a user’s origin, language, and a unique identifier. The server then selected whether to return the ordinary interface or an explicit one. A click on an associated Meta ad appeared to be one possible trigger.

The apps’ public faces included services for pet portraits, language learning, and AI workout coaching. The hidden interfaces WIRED described included image and video templates, image uploads, and purchases of in-app “diamonds.”

What the investigation connected

WIRED reported that the apps and websites shared code and references to other domains. Infoblox researcher Maël Le Touz identified almost 500 linked domains, a separate count from WIRED’s figure for more than 400 connected apps and domains.

WIRED also linked parts of the operation to Chinese-language developers and Hong Kong-registered KeLing Technology Limited, based on domain and service connections. KeLing did not respond to the publication’s request for comment.

What the platforms reportedly did

WIRED reported that Apple removed 19 apps after the publication contacted the company. Apple spokesperson Adam Dema said the apps had complied with Apple’s rules when first reviewed, then their developers changed the websites supplying the apps’ content after approval and promoted them through pornographic ads on a third-party platform.

WIRED said Google had removed most of the apps before the publication contacted the company. Meta told WIRED it had acted against 11,600 ad accounts, more than 4,700 business pages, 5,200 users, and 15,000 pages linked to the domains. Meta also said it blocked 209 apps from advertising on its platforms.

What Apple’s October 16 rules say

Apple published a separate App Review update on October 16, 2026. The notice explicitly includes nonconsensual intimate imagery, including imagery generated with AI, in content-moderation requirements. Apple says apps must proactively prevent and remove that material. Apple’s App Review notice also clarifies that apps that misrepresent their purpose during review and change after approval can be removed from the App Store and the Apple Developer Program.