Meta sues AI ‘nudify’ app Crush AI for advertising on its platforms

Meta has initiated legal proceedings against Crush AI, the developer of a controversial AI application known for generating nude images. The lawsuit, filed in Hong Kong, alleges that Joy Timeline HK, the organization behind Crush AI, attempted to bypass Meta's advertising review process to promote its nudifying services. In a recent blog update, Meta revealed that it had repeatedly removed ads from Crush AI for breaching its guidelines; however, the company claims that Joy Timeline HK persisted in placing new advertisements. According to Alexios Mantzarlis, the author of Faked Up, Crush AI reportedly ran over 8,000 ads for its 'AI undresser' services on Meta's platforms within just the first two weeks of 2025. Notably, Mantzarlis's research indicated that approximately 90% of Crush AI's website traffic originated from Facebook and Instagram, prompting him to alert Meta about several of these sites. Crush AI allegedly circumvented Meta's ad screening protocols by creating multiple advertiser accounts and frequently changing its domain names. Many of these accounts bore names resembling 'Eraser Anyone’s Clothes' followed by a series of numbers. At one stage, Crush AI even maintained a Facebook page to promote its services. This issue is not exclusive to Meta, as various social media platforms, including X, are grappling with similar challenges. As companies like Meta and TikTok strive to integrate generative AI into their services, they face the ongoing struggle of moderating tools that can potentially endanger users, particularly minors. A surge in links to AI nudifying apps was reported in 2024 across multiple platforms, including X, Reddit, and YouTube, where millions encountered advertisements for these applications. In response to this escalating issue, Meta and TikTok have prohibited keyword searches associated with AI nudifying apps, yet the complete removal of these services from their platforms remains a complex task. To combat this growing threat, Meta has developed new technologies aimed at identifying ads related to AI nudifying services, even when such ads do not display explicit content. The company is now employing matching technology to expedite the removal of copycat advertisements and has broadened the list of flagged terms, phrases, and emojis. Meta also stated that it is applying its established strategies for disrupting networks of malicious actors to tackle the new networks of accounts responsible for promoting AI nudifying ads. Since the beginning of 2025, Meta claims to have dismantled four distinct networks engaged in advertising these services. Beyond its platforms, Meta intends to share information about AI nudifying apps via the Tech Coalition's Lantern program, a collaborative initiative with other tech giants aimed at preventing child exploitation online. Since March, Meta has contributed over 3,800 unique URLs to this network. On the legislative front, Meta has expressed its commitment to supporting measures that empower parents to oversee their children’s app downloads, including backing the US Take It Down Act and collaborating with lawmakers to advance its implementation.