Meta Files Suit Against ‘Nudify’ App Crush AI Over Platform Ads

Meta has filed a lawsuit against the creator of Crush AI, a widely used AI “nudify” app, which allegedly ran thousands of advertisements on Meta’s platforms. Alongside the legal action, Meta announced new efforts to curb similar apps.

The lawsuit, filed in Hong Kong, accuses Joy Timeline HK—the company behind Crush AI—of trying to bypass Meta’s ad review system to promote its AI nudification services. According to a Meta blog post, the company repeatedly removed the ads for violating its policies, but Joy Timeline HK allegedly continued submitting new ads despite the removals.

Meta Ads Fueled Rise of AI-Powered Deepfake App ‘Crush AI,’ Report Reveals

Crush AI, an app that uses generative AI to create fake, sexually explicit images of real individuals without their consent, allegedly ran over 8,000 ads promoting its “AI undresser” services on Meta’s platforms in the first two weeks of 2025, according to Alexios Mantzarlis, author of the Faked Up newsletter. In a January report, Mantzarlis noted that around 90% of traffic to Crush AI’s websites originated from Facebook or Instagram, and that he had reported several of these sites to Meta.

To bypass Meta’s ad review system, Crush AI reportedly created dozens of advertiser accounts and frequently rotated domain names. Many of these accounts, according to Mantzarlis, were named “Eraser Annyone’s Clothes” with varying numbers. At one point, the company even operated a Facebook page to promote its services.

Facebook and Instagram aren’t alone in facing these issues. As platforms like X and Meta rush to integrate generative AI into their services, they’re also grappling with the difficulty of moderating how these tools may contribute to unsafe environments—especially for minors.

Platforms Struggle to Contain Surge of AI Undressing Apps Despite New Detection Efforts

Researchers observed a surge in links to AI undressing apps across platforms such as X, Reddit, and YouTube in 2024, with millions of users reportedly exposed to ads for these apps. In response, Meta and TikTok have blocked keyword searches related to AI nudify tools, though completely removing such content remains a major challenge.

Meta noted in a blog post that it has developed new detection technology to identify AI nudify or undressing ads—even those without explicit content. The company also said it’s employing matching technology to more efficiently locate and eliminate similar ads, and has broadened the range of keywords, phrases, and emojis flagged by its systems.

Meta stated that it is now using the same strategies it has historically employed to dismantle coordinated networks of malicious actors to target groups running ads for AI nudify services. Since the beginning of 2025, the company claims to have disrupted four such networks.

Beyond its own platforms, Meta also announced plans to share information about AI nudify apps through the Tech Coalition’s Lantern program—a joint initiative involving Google, Meta, Snap, and others aimed at combating child sexual exploitation online. Since March, Meta says it has contributed over 3,800 unique URLs to this collaborative effort.

On the policy side, Meta stated it will “continue to support legislation that gives parents the ability to manage and approve the apps their teens download.” The company previously backed the U.S. Take It Down Act and is currently collaborating with lawmakers on its implementation.

Read the original article on: TechCrunch

Read more: Meta’s V-JEPA 2 Model Trains AI To Understand Its Surroundings