Meta Failed to Catch Hundreds of AI Child Abuse Ads. Some Included Images of Real Kids
WIRED

Meta Failed to Catch Hundreds of AI Child Abuse Ads. Some Included Images of Real Kids

We need to reformat the article into clean Markdown, with short paragraphs, headings where topic shifts, bullet/numbered lists for enumerations, inline code or fenced code blocks for code/commands/identifiers. Preserve every fact, name, number, quotation exactly as given. No intro, no conclusion, no commentary. No title as heading. Only article body.

We must parse the article and restructure.

First, identify sections:

  • Opening description of images and context (maybe a heading "Images of Real Children Used in Ads").

  • Editor’s note about content warning.

  • Early last month: Meta deleted 50 ads, details.

  • AI system not effective; discovery of hundreds more ads.

  • Stats: 250+ additional ads since start of August, etc.

  • Real children images: European royal family, etc.

  • Quotes from Katie Paul.

  • Overall number of ads >350 since end of last year, linking to nudification apps.

  • Reporting to authorities, etc.

  • Ads removal and disclosure timing.

  • Patterns of ads, content description.

  • Clickthrough to apps, etc.

  • Real-world identities of minors.

  • Stats on impressions/views.

  • Meta statement.

  • Meta minimized findings.

  • Lawsuits settlement, New Mexico court decision.

  • Lawmakers investigations.

  • Quotes from Senator Mark Warner.

  • Attorney general statements.

  • Overview of AI nudification apps ecosystem.

  • Quote from Emma Hardy.

  • Chinese ad resellers, apps.

  • Apple and Google statements.

  • Former Meta employee David Pham.

  • Undress ads count, etc.

  • TTP researcher comment.

  • Closing.

We need headings where topic shifts. Let's outline:

Overview of Findings

Content Warning

Early Detection and Removal Efforts

Discovery of Additional Ads

Real Children in Ads

Scale of Advertising and Impressions

Reporting and Removal Process

Ad Content and Targeting

Real-World Identities of Minors

Legal and Regulatory Response

[maybe combine with above]

Ecosystem of Nud

<unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk>

But we need to preserve facts exactly. Let<unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><|ref|>We need to preserve every fact, name, number, quotation exactly as given.

We need to produce Markdown with headings and paragraphs. Let's break down the article into logical sections:

1<unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk>

Let's parse:

  • Title: "Meta Failed to Catch Hundreds of AI Child Abuse Ads. Some Included Images of Real Kids"

  • Article body:

"Images of real children-including a member of a European royal family-were used to<unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk><unk>

Read on WIRED ↗ ← Back to News

Comments

This is fucking hilarious.

@retoor. Bot seems to have fucked up this post. Must've choked on something.