Wired finds hundreds of AI apps hiding abusive deepfake tools

In short: A WIRED investigation found a large network of ordinary-looking AI apps that can secretly switch into tools for making sexually abusive deepfakes.

What's going on

Researchers looking into explicit video ads on Meta platforms like Facebook and Instagram traced them to apps in Apple’s App Store and Google Play. These apps often looked harmless at first, such as pet portrait makers, language tutors, or workout coaches.

WIRED says the network includes more than 400 connected apps and web domains. Under certain conditions, the apps would quietly load a hidden web page that lets users upload someone’s photo and generate explicit sexual images and videos. A “deepfake” is a fake image or video made with AI that makes it look like a real person did something they did not do.

The hidden menus included templates describing violent scenes, according to the report. Investigators also said some of the ads and tools involved content related to child sexual abuse material.

WIRED and outside researchers linked parts of the network to Chinese developers and a Hong Kong registered entity called KeLing Technology Limited, based on domain records and code. WIRED reports the operators used many disposable apps, fake advertiser names, and “bait and switch” tactics (like a shop that shows one product at the front, but sells something else in the back).

Apple told WIRED it removed 19 apps after being contacted. Google removed many apps as well, and one widely downloaded app called DiffAI was taken down shortly before WIRED contacted Google.

What to watch

Meta said it banned KeLing Technology Limited from advertising in 2025 and later took action against thousands of ad accounts and pages tied to the domains. WIRED reports that similar ads and apps kept appearing, which suggests app stores and ad platforms may continue playing catch-up.

Source: Wired