SAN FRANCISCO — City Attorney David Chiu has demanded that Meta stop accepting and distributing paid advertisements containing apparent AI-generated child sexual abuse material and explain how the ads passed through its review systems.
In a September 9 cease-and-desist letter to Meta Chief Legal Officer C.J. Mahoney, Chiu cited research identifying 332 apparent offending advertisements that ran on Facebook and Instagram. The letter says 298 directed users to AI image- or video-generation applications, while 182 pointed to applications researchers said could “nudify” people or place them into sexual videos.
The figures and descriptions come from the city attorney’s letter and the underlying investigation. They are allegations, not judicial findings.
An enforcement warning—not a court order
The letter asks Meta to immediately cease allowing paid advertisements containing child sexual abuse material and gives the company 28 days to provide an initial written response. Chiu’s office also requested information about Meta’s screening, escalation and reporting procedures, its handling of repeat offenders and advertising partners, and measures intended to prevent similar ads from returning.
This is not a lawsuit, court order or finding of liability. The city attorney reserved the right to conduct a further investigation and pursue legal remedies if Meta does not provide sufficient information or assurances.
The letter points to potential obligations under 18 U.S.C. § 2258A, which requires covered online service providers to report certain apparent child exploitation to the National Center for Missing & Exploited Children after obtaining the legally required knowledge. It also raises the possibility of enforcement under California’s unfair-competition law. The letter does not establish that Meta violated either law.
Meta told WIRED that it removed the advertisements identified by researchers and disputed San Francisco’s jurisdiction.
Why this matters to radio and media
The enforcement focus is significant because it moves beyond harmful user posts to a platform’s paid-advertising operation—where the platform reviews creative material, accepts money and controls distribution.
Radio stations and other media companies are not accused of wrongdoing in this matter. However, broadcasters that advertise AI applications or use third-party agencies for social campaigns should verify the services they promote, reject “nudify” and deceptive face-manipulation products, and require campaign partners to comply with child-safety and platform rules.
The practical question for media businesses is whether accepting advertising revenue creates a higher level of responsibility than merely hosting user content. Regulators are now examining whether review systems, reporting processes and reseller relationships are strong enough to keep prohibited AI-generated material out of paid distribution.
The RNN angle: Meta says advertisements are reviewed before distribution, yet the city attorney says hundreds of apparent child-exploitation ads entered the paid system. The next legal fight could center on what platforms knew, when they knew it and whether they fulfilled their reporting duties.
Read the city attorney’s September 9 letter.
This report provides general information and is not legal advice.