San Francisco's city attorney has formally ordered Meta to stop allowing adverts that use AI to transform images of children into sexually explicit videos, after investigative reporting found the company had run hundreds of such adverts on Facebook, Instagram and Threads. The adverts directed users to AI image- and video-generation apps capable of "digitally undressing" people or creating other non-consensual explicit imagery, with some targeting confirmed real minors, including a member of a European royal family. City attorney David Chiu accused Meta of profiting from a known problem it has failed to fix, calling the scale and persistence of the issue evidence that the company's safeguards are "woefully inadequate."

Researchers at the Tech Transparency Project identified more than 350 such adverts reaching over 29,000 accounts across the EU, US, Australia and India, with more than 250 appearing even after an initial WIRED report in early August flagged 53 of them. Meta says it has since removed all the adverts, noting most had fewer than 200 impressions and total spend was under $5,000, and disputes San Francisco's jurisdiction by claiming no evidence the ads were shown in the city. The City Attorney's Office rejected that defence, arguing residents could still access the adverts online, and has sent Meta a four-page letter demanding explanations of its moderation failures, its reporting practices to the National Center for Missing and Exploited Children, and how it handles repeat offenders.