Apple has removed 19 apps from the App Store after an investigation found they were part of a network of hundreds of apps that hid tools for generating nonconsensual intimate images of real people behind ordinary-looking services. The findings, published on November 2 by the nonprofit AI Forensics and reported by WIRED, identified 444 connected apps and web domains that presented themselves as pet portrait editors, language tutors, fitness coaches and similar everyday tools.

How the switch worked

According to the researchers, the apps contacted remote servers after installation, and the server decided which version of the app a user would see. Reviewers and ordinary users could be shown the advertised feature, while other users were routed to a separate web page offering the hidden tools. One apparent trigger was installing an app after tapping an advertisement on Facebook or Instagram. One example cited in the reporting was Viefur, an app presented as a pet portrait generator, which Apple later removed.

The scale of the network was large. The researchers counted more than 84,000 advertisements on Meta's platforms tied to the network since January 1, paid for under fictitious payer names. As of September, 27 related apps were still live on Google Play with more than 350,000 downloads between them, and 21 were on the App Store, according to the report. A separate researcher identified almost 500 linked domains running the same code. Some of the apps carried age ratings indicating they were suitable for all ages. WIRED linked the operation to KeLing Technology Limited, a company registered in Hong Kong.

What the platforms said

Apple said the apps complied with its rules when they were reviewed, but that their developers later changed the website content the apps loaded and promoted them through advertisements on a third-party platform. After WIRED shared the domains, Apple removed the 19 apps and said it had suspended tens of thousands of apps over the past year for similar bait-and-switch tactics. The company had already tightened its App Store rules on October 16, with Guideline 1.2 addressing AI-generated nonconsensual intimate imagery and Guideline 2.5.19 prohibiting hidden or undocumented features activated after approval.

Google had already removed most of the related apps before WIRED contacted it and did not comment. Meta said it had banned KeLing from advertising in 2025 and has since acted against 11,600 advertising accounts and 15,000 pages linked to the domains, and blocked 209 apps from advertising.

Why review alone missed it

The case exposes a limitation of reviewing apps before publication: content delivered from a developer's own servers can change without a new version of the app being submitted for approval, allowing prohibited features to appear after an app has passed review. The researchers said similar apps were still appearing even as the reporting was completed, suggesting the network's removal is not the end of the problem.

For app stores, the episode is a test of whether the tougher rules adopted last month can keep pace with operators who design their apps to look harmless until after they are installed. The platforms say enforcement is continuing; the investigation suggests reviewers, advertisers and stores will all have to watch what happens after approval, not just before it.