Apps presented as everyday AI tools concealed abusive deepfake features, WIRED reported November 2.

The big change

  • WIRED identified over 400 connected apps and domains.
  • A server selected the interface.
  • Apple’s rules prohibit hidden app features.

How the concealment worked

AI Forensics researcher Paul Bouchaud told WIRED that servers selected benign or abusive web interfaces. The criteria were unclear; downloads following associated Meta ads appeared to be one trigger. WIRED reported that the hidden interfaces let users upload people’s photos into sexually explicit image and video templates.

BIG CHANGE did not independently test the apps. The linked AI Forensics report was unavailable when checked.

Apple’s rules and response

Apple removed 19 apps after notification, telling WIRED their web content changed after review. WIRED said Google had already removed most identified apps and did not comment. Meta told WIRED it took action against thousands of associated accounts and pages.

Separately, Apple’s October 16 policy update says apps must proactively prevent and remove nonconsensual intimate imagery, including AI-generated content. It also clarifies that apps misrepresenting functionality during review and changing after approval face removal from the store and developer program.

BIG CHANGE previously covered Apple’s separate rule changes.

Sources & further reading