San Francisco targets Meta over AI-generated child abuse ads

San Francisco’s City Attorney’s Office has sent Meta a four-page cease-and-desist letter over paid advertisements containing AI-generated child sexual abuse material. The letter, reportedly sent on 9 September 2026, asks Meta to stop allowing such advertisements on Facebook, Instagram and Threads and to provide information about its moderation and reporting processes.

The action follows findings by the Tech Transparency Project (TTP), which identified more than 350 advertisements in recent months that used images of minors, including a recognised European royal, to create short videos depicting sexual acts. The advertisements linked to AI image and video generation applications capable of generating sexualised images of individuals.

According to TTP, the advertisements were shown to more than 29,000 accounts across the European Union and also reached users in the United States, Australia and India.

San Francisco City Attorney David Chiu said Meta had failed to address a known issue involving AI-generated child sexual abuse advertisements while profiting from them. The cease-and-desist letter states that the findings point to broader problems with Meta’s ability to detect child sexual exploitation in advertising and raise legal questions concerning the company.

The letter asks Meta to explain how the advertisements repeatedly passed through its advertising systems and to engage with city officials on child safety, advertising and legal concerns.

Meta told WIRED that there was no indication the advertisements had been shown in San Francisco and argued that the matter fell outside the City Attorney’s jurisdiction. The company said the identified advertisements had been removed for violating its policies and reiterated that it does not tolerate child exploitation or applications designed to create sexualised images of individuals.

Meta also said that total spending across the more than 300 advertisements was below $5,000. However, TTP director Katie Paul told WIRED that new advertisements featuring the same children continued to appear, arguing that the removals had not fully stopped the activity.

The city’s letter gives Meta 28 days to provide information on how the advertisements evaded moderation, how the company escalates and reports such material to the National Center for Missing & Exploited Children (NCMEC), and how it addresses repeat offenders.