Meta Ran More Than 350 AI Child Abuse Ads Since Late 2025, Reaching 29,000 People, Researchers Find

Meta says every ad on its platforms gets reviewed before it runs. The numbers coming out of a nonprofit watchdog contradict that claim.

The Tech Transparency Project, a research arm of the Campaign for Accountability, reported that Meta has run more than 350 AI-generated video ads depicting child sexual abuse across Facebook, Instagram and Threads since the end of last year, according to Bloomberg. TTP says the ads collectively reached more than 29,000 people.

This isn't the first time TTP caught this. In August, following reporting by WIRED, Meta pulled roughly 50 ads containing child sexual abuse material after researchers flagged them. Meta told WIRED at the time that most of those ads predated a new AI detection system the company had rolled out specifically to catch this kind of content.

TTP told WIRED it found more than 250 additional CSAM ads running on Meta's platforms since the start of August, some identical to the ones Meta had just taken down.

The new batch included something the first round didn't: real children. TTP identified ads that used an officially released photograph of a minor from a European royal family and turned it into a video depicting a "graphic sex act," according to WIRED. Researchers aren't naming the royal to protect the child and say they've alerted officials in the country. WIRED independently confirmed the source photo came from the royal family's official website.

"There's a lot of harm that can come from this all while Meta is raking in money in the process," TTP director Katie Paul told WIRED.

Most of the roughly 300 related ads TTP tracked pointed users toward AI "nudify" apps, many tied to China-based developers, according to WIRED's reporting. Apple spokesperson Adam Dema told WIRED the company has "zero tolerance for developers who deliberately attempt to evade our review process" and said it had already pulled 17 of 20 flagged apps, taking action on the rest, plus seven more WIRED identified separately. Google spokesperson Dan Jackson said the company doesn't allow nudify apps, has suspended hundreds of violators, and restricted search terms like "nudify" on the Play Store.

David Pham, a former Meta employee who investigated nonconsensual imagery on the platform before leaving last year, told WIRED that people posting this content adapt almost instantly. "Every time the platform or investigators would build a new detection model or a new way to have some type of enforcement, they quickly, within a day, maybe five days, built a new way to avoid detection," Pham said. He added that Meta has the resources to do better: "I would like to live in a world where we're not getting CSAM or NCII material sent directly to our phones."

Warner Wants Answers, Not Excuses

Sen. Mark Warner (D-VA) sent a letter to Meta CEO Mark Zuckerberg demanding an explanation, according to a statement from his office. Warner pointed out that Meta's own ad standards prohibit sexually suggestive content and that the company says it reviews every ad before it runs.

"It strains credulity that Meta was (or is) unknowingly hosting, and profiting from, such blatantly illegal content, when your company explicitly states that it reviews and approves all ads," Warner wrote. He asked Meta how the ads got approved, how much money the company made off them, and what it's doing to stop new ones from appearing.

Warner's letter also noted that a July 2026 BBC investigation had already found similar CSAM-linked ads running on Instagram in India, and that new ads kept surfacing even after Meta was directly notified. No congressional investigation into Meta has been announced, and Meta has not publicly responded to Warner's letter as of this writing.

Warner has been pushing on this front for a while. In July he introduced the SAFE AI Act, which would bar federal agencies from buying AI systems capable of generating CSAM or non-consensual intimate images and would give survivors a legal path to sue over AI-generated abuse material.

A Separate, Thornier Legal Fight

While Meta deals with the ad problem, a different fight is playing out in federal court over what counts as illegal in the first place. In late August, a three-judge panel on the U.S. Court of Appeals for the 7th Circuit tossed a possession charge against Steven Anderegg, a software engineer who used AI tools to generate what prosecutors called "hyper-realistic" CSAM.

The panel ruled that under Supreme Court precedent, private possession of AI-generated CSAM that doesn't depict an actual child is protected by the First Amendment, citing the 1969 case Stanley v. Georgia. Anderegg still faces separate charges for producing and distributing the material, and for sending an image to a minor over Instagram, a message Meta itself detected and reported to the National Center for Missing and Exploited Children.

Dr. Mary Anne Franks, a law professor at George Washington University, told Mashable she understands why the ruling sounds alarming but called it a straightforward application of decades-old precedent. Notably, the panel itself flagged discomfort with its own conclusion, writing that it would "benefit from additional guidance from the Supreme Court regarding the intersection between the First Amendment and virtual CSAM if an appropriate case should arise."

Whether that case reaches the Supreme Court is an open question. So is whether Meta, facing a senator's letter and a stack of youth-safety lawsuits already numbering more than 3,000 against Meta, TikTok, Google and Snap, will actually close the gap between what its ad policy claims and what its ad system lets through.