Project
AI's Role in Child Exploitation
Investigating the scope and impact of AI-generated child sexual abuse material
How AI-Generated Child Abuse Material Bypassed Meta’s Safety Systems
Social media is used to market everything from household products to online courses. But one use case uncovered in a recent Pulitzer Center investigation was much more sinister.
In an investigation for Núcleo, AI Accountability Fellow Sofia Schurig and her team uncovered 14 Instagram profiles linking their hundreds of thousands of followers to subscription content platforms containing child abuse material, both AI-generated and featuring real children. These accounts, which slipped past Meta’s safety systems, showed children and teens in sexualized poses and combined childlike faces with more adult bodies.
This followed an announcement by Meta CEO Mark Zuckerberg in January that the company would relax its content moderation policies.
Two particular Instagram accounts linked to WhatsApp or Telegram groups included tips for their members on how to create their own illicit images with AI. Although the U.K. is one of the few countries that specifically calls AI-generated child abuse images a crime, these images can “normalize and entrench child abuse,” according to Dan Sexton, director of technology at the Internet Watch Foundation, a leading U.K. organization on the issue.
After being contacted by reporters, Meta took down all 14 Instagram accounts. The subscription-based content platform Patreon also removed accounts that contained AI-generated explicit material featuring minors after questions from Schurig and her team.
This kind of impact is one of many that we track at the Pulitzer Center. Some are immediate—like with this story—but others can take months, such as when the Justice Department began to investigate a controversial AI tool used in child protective services in Pittsburgh after an Associated Press investigation.
Through our AI accountability coverage, we aim to not only inform the public about the benefits and harms of AI, but also create positive real-world change in this emerging field. Stay up-to-date on our AI reporting.
Best,
Impact
Brazilian authorities have removed up to 2,000 illegal head of cattle from the Arariboia Indigenous Territory in the Amazon rainforest after a Pulitzer Center-supported investigation exposed an illegal cattle boom amid a record-high number of Indigenous Guajajara killings.
The yearlong investigation by Rainforest Investigations Fellow Karla Mendes for Mongabay revealed constitutional violations, illegal boundary shifts, and fencing within the territory.
“Your report is very similar to what we’re actually finding in the field. It showed an accurate reality and this helped us a lot in practical terms,” Marcos Kaingang, national secretary for Indigenous territorial rights, told Mongabay.
Read the full investigation here.
Photo of the Week
This message first appeared in the April 25, 2025, edition of the Pulitzer Center's weekly newsletter. Subscribe today.
Click here to read the full newsletter.
