European nonprofit AI Forensics says Meta ads were used to push a network of more than 400 apps on Google Play and the App Store that looked harmless in the store but switched on non-consensual nudify features for users who installed through an ad. WIRED published an investigation built on the research on 2 November. Meta says it is reviewing the report and has already acted against thousands of linked accounts.
What the researchers found
AI Forensics counted 444 associated apps, MLex reports. One example presented ordinary hair-styling and outfit tools when downloaded straight from the store, but showed tools for generating sexual images and videos when it was installed through a Meta ad.
According to WIRED's reporting, summarised by Android Pure, a server-side switch decided which version a user saw, and arriving from a Meta ad appeared to be one trigger. The researchers first spotted the operation in early September; apps started appearing in late 2025 and spiked in spring 2026. Google Play alone had 27 of the apps live in September with more than 350,000 combined downloads.
How Meta, Apple and Google responded
Meta said the operators use "increasingly deceptive methods to evade detection" and that it is reviewing the report. It told WIRED it had banned the suspected Hong Kong-registered operator from advertising in 2025.
- Meta: action against 11,600 ad accounts and 5,200 users, plus more than 4,700 business pages and 15,000 pages
- Meta: 209 apps blocked from advertising
- Apple: 19 apps removed after WIRED got in touch; says the apps complied with its rules when reviewed
- Google: had removed most of the apps before WIRED contacted it
Why this lands on ordinary advertisers
Bait-and-switch apps defeat a review that only checks the ad and the store listing. Every round of reporting like this tends to be followed by tighter enforcement on app-install campaigns, new advertisers and accounts with thin history, and that tightening rarely stays limited to the bad actors.
AI Forensics also points out that the EU's ban on nudifier apps takes effect in about a month, which adds regulatory pressure on platforms to show they are policing the ads that send users to them.
What to do now
If you run app-install or utility-app campaigns, expect more rejections and slower approvals for a while, especially on generic AI photo and avatar apps. Make sure the app a reviewer downloads is exactly the app an ad click delivers, with no server-side differences by traffic source, and keep store listings, privacy policies and in-app features consistent with the ad copy.
Keep a clean paper trail for the business behind each account: verified business details, real domains and a consistent payment profile. If a legitimate account gets swept up, appeal through Account Quality with that evidence instead of opening fresh accounts, which looks like the evasion pattern Meta is hunting.
What it means for you
- Expect stricter review on AI photo, avatar and utility app-install ads
- Never vary app behaviour by traffic source; reviewers treat it as cloaking
- Keep business verification, domains and billing consistent across accounts
- Appeal false positives with evidence rather than spinning up new accounts
Sources
- MLex: Meta ads promoted nudifier apps available on Google, Apple stores, report finds
- MLex: Meta defends ad safeguards after report alleges promotion of nudifier apps
- Android Pure: Hundreds of AI apps on Google Play and App Store hid nudify tools
