San Francisco City Attorney Targets Meta Over Proliferation of AI-Generated Child Sexual Abuse Material in Paid Advertisements

San Francisco City Attorney David Chiu has formally issued a cease-and-desist letter to Meta Platforms Inc., demanding an immediate end to the display of paid advertisements containing AI-generated child sexual abuse imagery on the company’s social media platforms. The legal action follows a series of investigative reports revealing that Meta’s advertising infrastructure has been utilized to circulate hundreds of videos and images depicting minors—including, in some instances, real individuals—engaged in nonconsensual sexual acts.

The controversy centers on the Tech Transparency Project (TTP), a research organization that identified more than 350 instances of these advertisements running on Facebook, Instagram, and Threads in recent months. These ads function as a gateway to external applications capable of performing "nudification"—the digital manipulation of images to remove clothing or simulate sexual activity. The persistence of these ads has drawn sharp criticism from local government officials, who argue that Meta’s failure to prevent the monetization of such content represents a profound systemic breakdown in its safety and moderation protocols.

A Pattern of Escalating Failures

The chronology of these findings suggests that Meta’s internal safeguards were aware of, yet failed to effectively mitigate, the problem. In early August, initial reports surfaced identifying 53 specific advertisements that violated safety policies. Despite this public notice, the issue not only continued but expanded. Researchers discovered that in the weeks following that first report, more than 250 additional ads appeared on Meta platforms.

The scale of the distribution is significant. TTP data indicates that these ads collectively reached more than 29,000 accounts across the European Union, with additional exposure reported in the United States, Australia, and India. The ads were not merely ephemeral; many were re-uploaded multiple times, suggesting that automated moderation systems were unable to recognize previously flagged content. In several instances, the ads utilized still images of real children—including a member of a European royal family—transforming them into short, sexually explicit video clips through generative AI tools.

The Legal and Ethical Standoff

In his four-page letter to Meta’s legal counsel, David Chiu emphasized the hypocrisy of the company’s business model. "No company—particularly one that claims all advertisements are reviewed and approved prior to distribution and then accepts payment for that distribution—should permit its advertising systems to be used in this manner," Chiu wrote.

Meta’s initial response to the accusations has centered on jurisdictional arguments. The company claimed there was "no indication" that the ads were displayed specifically to users in San Francisco, suggesting the matter fell outside the City Attorney’s jurisdiction. This defense was met with immediate, sharp rebuke from the City Attorney’s office.

"We find Meta’s response astonishing," said Alex Barrett-Shorter, a spokesperson for the office. "Meta appears less concerned about whether they’re guilty of exploiting children than whether San Francisco should say anything about it. San Francisco consumers can access the unlawful ads through Meta online spaces; therefore, this is the business of our city. Meta has bigger problems to solve here than determining our jurisdiction."

Behind the Numbers: The Economics of Exploitation

Meta has attempted to frame the issue as a minor oversight, telling journalists that the total ad spend for the 350-plus ads identified by TTP was under $5,000. The company maintained that most of these ads received "fewer than 200 impressions." However, safety advocates argue that focusing on the monetary value or the reach of individual ads misses the point of the broader regulatory failure.

The core concern is that Meta, a multi-billion-dollar enterprise with industry-leading AI detection capabilities, has allowed its paid advertising ecosystem to facilitate child sexual exploitation. When these ads are clicked, they direct users to third-party apps designed specifically for the sexual objectification of minors and adults. By allowing these apps to purchase ad space, Meta is effectively acting as an intermediary, generating profit from content that violates its own stated community standards.

Systemic Issues in Moderation

The TTP investigation highlights a worrying lag in Meta’s response time. In several cases, researchers reported the advertisements to Meta, only to receive no action for over a week. During this delay, the advertisements continued to serve impressions to thousands of users. Furthermore, a review of Meta’s own ad library revealed that even when some ads were eventually removed, they were not consistently flagged or categorized as containing child sexual abuse material, which hinders the company’s ability to track repeat offenders and prevent future uploads.

Katie Paul, director of the Tech Transparency Project, noted that as of late September, new ads were still appearing on the platform featuring the same children previously reported to the company. "Meta has claimed it takes a hard stance against child sexual abuse material, but since our report was published, more brand-new ads have run," Paul stated. "The findings are clear: Meta is not effectively addressing this issue."

Broader Implications for the Tech Industry

The situation poses significant legal and reputational risks for Meta. The City Attorney’s demand for "prompt discussions" with Meta’s legal and child-safety teams aims to force the company to be transparent about its moderation pipeline. Specifically, the City is demanding details on:

1. How these ads bypassed existing automated and manual review systems.
2. The internal process for escalating identified child sexual abuse material to the National Center for Missing and Exploited Children (NCMEC).
3. The protocols for banning advertisers and blocking repeat offenders from purchasing future ad space.

The incident also raises questions about the efficacy of current legislation regarding platform liability. While Section 230 of the Communications Decency Act has historically provided a shield for platforms regarding third-party content, the fact that Meta is actively reviewing, approving, and accepting payment for these specific advertisements complicates the legal landscape. If a platform is paid to distribute content, it may be subject to stricter liability standards than if that content were merely user-generated posts.

Conclusion: A Crisis of Trust

Meta’s struggle to contain the rise of "nudify" apps and AI-generated exploitation is not new. The company has previously sued entities associated with these apps and has publicly committed to working with other technology firms to curb the trend. Yet, the evidence presented by the San Francisco City Attorney and the TTP suggests that these efforts remain fragmented and reactive.

The demand for a response within 28 days places the onus on Meta to demonstrate that it has the technical and administrative willpower to rectify these failures. As AI technology continues to lower the barrier to entry for the creation of nonconsensual intimate imagery, the pressure on major social media platforms to implement fail-safe moderation will only increase. For San Francisco officials, the objective is clear: they are seeking not just the removal of current ads, but a fundamental overhaul of the systems that allowed such content to be monetized in the first place. Whether Meta can provide the transparency required to restore public trust remains an open, and increasingly urgent, question.