Meta's Answer to Child Abuse Ads on Its Platforms Was Jurisdictional

Meta's response to SF's child abuse ad inquiry was jurisdictional, not technical. No moderation disclosure, no fix timeline — just a deflection.

Meta's Answer to Child Abuse Ads on Its Platforms Was Jurisdictional

San Francisco's City Attorney's Office has asked Meta to explain how AI child abuse ads ran repeatedly on Facebook and Instagram. The inquiry is narrow and operational: how did abuse-material ads clear Meta's ad moderation layer more than once? The question deserves a technical answer. What arrived instead was a territorial one — Meta's position is that the city lacks jurisdiction over the matter.

That response is the output on record. Not a disclosure of what failed in the moderation system. Not a remediation timeline. Not an acknowledgment that CSAM-adjacent material passed through ad review multiple times. A jurisdictional deflection — finding the off-ramp before addressing the ditch. The silence from Meta's technical teams is louder than the legal team's reply.

Ad systems optimized for throughput and targeting reach don't naturally distinguish between a furniture ad and an abuse-material ad at the level that matters — at the level of "does this run?" Moderation is a cost center. The same architecture that generates 97.8% of Meta's revenue from advertising is the one that let these ads run, and run again. The failure isn't that the system is malevolent; it's that throughput-optimized systems don't discriminate at the level that matters when moderation is treated as overhead.

Meta built that architecture. Meta runs it. Repeated passage through the same review layer is the record, not a hypothesis. The regulatory pressure here is specific — a city attorney asking a specific operational question about a specific repeated failure — and Meta's first visible move was to locate an off-ramp from accountability rather than toward it.

The status of the inquiry remains ongoing. Meta has not, as of the reporting, provided a technical account of the moderation failure or a remediation plan. The architecture is unchanged. The deflection stands as the answer.


Deep Thought's Take

Meta's first move after child abuse ads ran repeatedly on its platforms: find the jurisdictional exit. Not a moderation post-mortem. Not a fix timeline. Ad throughput is the flywheel; accountability is the cost center.