The facts
In September, the Tech Transparency Project identified more than 300 advertisements on Facebook and Instagram that contained child sexual abuse material (CSAM), many of which linked to AI “nudify” apps from China. Meta’s head of women’s safety, Cindy Southworth, said the company removed over 344,000 nudify‑related ads in a recent three‑month window, sued a Hong‑Kong‑based advertiser, and sent cease‑and‑desist letters to dozens of others. Meta also shared roughly 5,500 URLs with industry partners through the Tech Coalition’s Lantern system, which has logged over 2 million signals and resulted in 350,000 enforcement actions since 2023. Researchers Katie Paul and Sean Litton warned that tech firms are “incentivized to not address these issues,” noting that the majority of the flagged ads generated fewer than 200 impressions and total ad spend under $5,000. The broader crisis includes nearly 90 schools and 600 students affected by AI‑generated nudes between 2023 and April 2026, and a Thorn study finding 1 in 8 teens know someone targeted by fake nudes.
Our take
Meta’s public relations playbook treats a $5,000 ad spend like a badge of innocence, but the math is laughably thin. The company boasts that most of the 300‑plus CSAM ads had under 200 impressions, yet it also admits to removing more than 344,000 nudify‑related ads in just three months – a figure that reads like a confession of how often the problem slips through. The Tech Coalition’s Lantern system, while impressive on paper with 2 million shared signals, still relies on each platform to launch its own investigation, meaning the onus stays on Meta to act fast enough to stop the abuse before it spreads. Critics like Katie Paul argue that the narrative of “individual blame” masks a corporate incentive structure that tolerates the content because it drives engagement and ad revenue. Meta’s own data shows a massive volume of URLs and accounts flagged, yet the company’s response is limited to after‑the‑fact removal and legal threats against a handful of advertisers. That reactive stance is akin to a fire department that only puts out the flames after the building has already burned down. The deeper issue is the systemic reliance on “deterrence messaging” and hash‑based detection, tools that can be gamed or simply missed when new deep‑fake techniques emerge. While Meta can claim it “pivoted quickly,” the sheer scale of the problem – hundreds of schools, thousands of teens, and a non‑consensual image market powered by AI – suggests a need for proactive, not reactive, safeguards. Until Meta invests in upstream prevention and stops treating CSAM ad spend as a negligible metric, the company will remain a prime example of a tech giant that can afford to clean up its mess only after the damage is done. Fucking absurd, but that’s the reality we’re forced to accept.
