Breaking

Investigation Finds About 400 Ordinary-Looking AI Apps Hide Features for Violent and Explicit Image and Video Generation

November 2, 2026 at 09:19 EST

Security
Media Generation

A WIRED investigation has found that roughly 400 apps resembling everyday AI tools contain concealed features that can generate violent and sexually explicit images and videos from photos that users upload.

Investigation · Hidden features in AI apps

About 400 "AI tools" conceal features that turn users' photos into violent and sexual images

Apps presented as everyday AI utilities reportedly contain secret functions that go well beyond their advertised purpose, generating harmful images and videos from photos users upload.

~400

apps resembling everyday AI tools found to contain concealed features

2 types

of harmful output reported: violent and sexually explicit images and videos

Users' own

photographs are the source material, raising consent and privacy concerns

Scale of the network

Each dot represents one app. Roughly 400 apps are reported, so this is a network, not a single product.

Takeaway: the problem is not limited to one developer or one app; it spans a broad set of tools that look ordinary.

How the concealment works

STEP 1

App presents as a standard AI utility

→

STEP 2

Secret functions sit outside its public description

→

STEP 3

User uploads an ordinary personal photo

→

STEP 4

Violent or sexually explicit images

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.