Meta Ran Paid Ads with AI-Generated Child Sexual Abuse Imagery — and India Demanded Answers

Over a nine-month period, Meta approved and ran paid advertisements on its platforms that explicitly featured AI-generated child sexual abuse material (CSAM). WIRED reported that "dozens of paid ads" containing such imagery appeared on Facebook, Instagram, Threads, and Messenger, with some remaining active even after Meta was directly alerted .

The Tech Transparency Project (TTP) independently identified more than 50 of these image and video ads inside Meta's own public Ad Library, posted between November 2025 and August 2026. Several of the ads linked to so-called "nudify" apps that generate non-consensual sexualized images . The ads collectively reached thousands of accounts across the US, UK, and more than a dozen European countries .

How Meta's Own Systems Failed

Meta's advertising standards explicitly forbid ads that contain CSAM or promote nudify apps . Yet the ads were discovered using Meta's own transparency tools. TTP researchers found that some ads were linked to Beijing-based GatherOne Inc., an approved Meta advertising partner, which had placed approximately 7,600 ads for AI nudify apps across Facebook and Instagram . TTP identified 210 Facebook pages tied to GatherOne, with 85 of those pages still active at the time of reporting .

Meta later stated that most of the ads had minimal reach and were disabled after review, attributing the failure to detection systems that predate its newer AI moderation tools .

India's Swift Regulatory Response

The Indian government reacted aggressively to the revelations. Following a separate BBC Eye investigation that found Instagram running paid ads explicitly promoting CSAM — using search terms like "rape video" and "child video" that linked to Telegram channels — India's Ministry of Electronics and Information Technology (MeitY) took multiple actions:

• IT Minister Ashwini Vaishnaw directed MeitY to summon Meta for a formal explanation .
• MeitY issued a formal notice ordering Meta to immediately disable all advertisements and content on Instagram that promote or facilitate access to Child Sexual Exploitative and Abuse Material (CSEAM) .
• Meta was given seven days to provide a detailed explanation or face legal action, including potential penalties under India's IT Act and POCSO Act .
• The government specifically demanded that Meta explain how its ad review system had passed such content for paid distribution .

The European Union's Forward-Looking Ban

The EU's regulatory response was structural and legislative rather than a specific enforcement action against Meta for these ads. The EU AI Act now includes a new provision that will forbid the generation of CSAM outright starting December 2, 2026 . This directly targets AI-generated abuse material like that in the WIRED report.

The AI Act also imposes obligations on deployers and providers of AI tools to ensure safety and transparency, and requires AI-generated content to be identifiable . Separately, the European Commission had already fined Meta €200 million in April 2025 for non-compliance with the Digital Markets Act over its "Consent or Pay" advertising model, though that action was not directly related to CSAM .

What This Means

The twin investigations reveal a systemic failure in Meta's ad review infrastructure. Despite clear policies, AI-generated CSAM slipped through for months across multiple platforms. India's rapid response — a 7-day ultimatum — signals growing impatience among governments with platform self-regulation. The EU's ban, meanwhile, sets a global precedent by outlawing the generation of CSAM itself, not just its distribution.