Skip to content
IFAI

Industry

Meta sues nudify app developer, ramps up crackdown

Published Reach 42

Meta just filed a lawsuit in Hong Kong against Joy Timeline HK Limited, the company behind the CrushAI nudify app, for repeatedly bypassing ad review rules. The company also rolled out new detection tech targeting benign-looking bait ads and shared over 3,800 violating URLs across the tech industry via the Lantern network.

Meta sues nudify app developer, ramps up crackdown
Image: about.fb.com

Meta is going to war against the flood of AI nudify tools polluting its ad feeds.

The tech giant filed a lawsuit in Hong Kong against Joy Timeline HK Limited, the developer behind the controversial CrushAI apps. The lawsuit aims to permanently ban the company from buying ads across Meta platforms after it repeatedly tried to bypass ad filters.

Meta banned ads promoting non-consensual deepfake nude tools over a year ago. But the operators behind them kept finding crafty ways back into users' feeds.

Catching ads that hide the nude pixels

These bad actors have turned evasion into a science.

Because standard nudity filters flag explicit pictures, app developers began running completely clean-looking imagery in their ad creatives. Once a landing page URL got blacklisted by Meta, they would spin up fresh domains in minutes.

To fight back, Meta built new detection systems specifically trained to spot nudify promotions even when the ads themselves show no nudity. The platform also updated its matching systems to squash copycat campaigns faster.

Meta expanded its blocklist to include subtle code phrases and suggestive emojis, and put its coordinated inauthentic behavior teams on the case. Since the start of the year, Meta has wiped out four separate ad-running account networks.

Meta has also blocked search queries like nudify, undress, and delete clothing on Facebook and Instagram. But since these apps live all over the web, Meta took the fight cross-platform.

Through the Tech Coalition’s Lantern program, Meta began feeding intelligence to rival tech firms. Since late March, it has shared over 3,800 unique offending URLs so other platforms can take action too. On the policy front, Meta is backing the federal TAKE IT DOWN Act and lobbying for rules giving parents more control over teen app downloads.

Why it matters

AI-powered non-consensual imagery is one of the most toxic corners of the generative AI boom, turning personal photos into deepfake explicit images with zero permission. For the shady developers running these tools, social media ad platforms are the primary customer acquisition funnel.

Meta’s multi-pronged approach here looks genuinely sharp. Simple keyword bans and pixel filters do almost nothing against financially motivated bad actors who churn through proxy domains all day. Tracking the coordinated account networks behind them and hauling a major operator into court hits them where it hurts.

Sharing 3,800 bad links with competitors is arguably the most impactful tactical move. If the entire tech ecosystem shares threat intelligence, rotating domains stops being an easy cheat code.

Still, we think Meta's crackdown highlights a glaring bottleneck in the supply chain: mobile app stores. As long as nudify utilities can slip onto app marketplaces or operate freely on decentralized web infrastructure, social platforms are just playing high-tech whack-a-mole. Pushing for legislative action shows Meta knows it cannot win this battle completely alone.

Sources (1)

  1. [1]
    about.fb.com(opens in a new tab)Originalabout.fb.com

Best of AI, daily