Journalists' inquiries prompted Meta to remove eight Facebook accounts that users had already reported for publishing artificial intelligence-generated depictions of athlete abuse.
The accounts had remained online despite those reports, raising questions about the company's handling of inappropriate content.
Here's what to know
Reporter Maggie Harrison Dupré's team used an ordinary personal Facebook account to flag 10 pages sharing AI-generated videos of older male coaches inappropriately touching young female athletes. She described the investigation in an Instagram video posted by Futurism.
Facebook cleared eight of the accounts a week after the reports were submitted, sending notifications that it had found no community standards violations and would leave them active. The other two accounts had disappeared, but the team could not determine why.
Many of the athletes looked like teenagers, Harrison Dupré said. She described clips lasting just a few seconds that showed unwanted contact, grooming, and sexual abuse.
One Instagram user expressed frustration with the company's response.
TCD Picks » Quince Spotlight
💡These best-sellers from Quince deliver affordable, sustainable luxury for all
"I don't think Meta has strong moderation when it comes to inappropriate content. Who knows if it's AI or keyword moderated or if it even gets escalated to an actual human. I've reported very direct and obvious self-harm comments, and it comes back, 'This content doesn't violate our community standards,'" they said.
More background
An expert Harrison Dupré consulted agreed with her assessment of the videos as depictions of sexual abuse.
Harrison Dupré also identified herself as a Division I athlete.
She said the videos presented abusive conduct as normal by showing the young athletes responding to their coaches with smiles, laughter, or thanks.
New pages sharing similar videos appeared every day, Harrison Dupré said. During the investigation, Facebook filled her recommendations with this content from additional accounts.
Harrison Dupré questioned Facebook's failure to remove reported accounts while it continued recommending that kind of material.
Beyond the most obvious moral implications, the technology behind these AI-generated videos also raises environmental concerns. The computing infrastructure that supports AI requires electricity and can consume substantial amounts of water for cooling.
On the other hand, AI could also help optimize clean energy systems.
What's being done?
Meta removed the remaining pages following direct contact from Futurism's journalists, who supplied the list of flagged accounts after the initial reporting process failed.
Updated notifications reversed several earlier findings, stating that the reported material violated community standards and that the accounts had been deactivated.
Harrison Dupré said Meta's initial response emphasized that its platform had no place for abuse. Pressed to explain the differing outcomes, the company acknowledged mistakes but maintained that it moderated proactively.
Harrison Dupré continued to see additional pages in her recommendations even after Meta deleted the accounts the journalists had identified.
She said account-by-account enforcement could not easily keep pace with publishers' ability to generate and post similar videos again.
"And because at the end of the day, a journalist flagging 10 specific accounts to Meta and then deleting 10 specific accounts is not content moderation. Because of AI, we really need platforms to be proactive about abusive content, about abusive material. And so, you know, to see this is just really disappointing," Dupré said.
Where can I learn more?
Concerns about AI-generated content on social platforms extend beyond the athlete videos. These stories examine manipulative videos, fake fashion clips, and joke theft.
• On Meta's platforms, AI-generated videos of crying children solicit likes by promising their release from arrest.
• On TikTok, crying AI-generated women hide GLP-1 injections in videos repeatedly promoting the same app.
• Fashion fans mistook AI-generated outfit videos for real footage before discovering the deception.
• On Instagram, fake AI comedians are stealing real comedians' jokes without an effective platform fix.
Get TCD's free newsletters for easy tips, smart advice, and a chance to earn $5,000 toward home upgrades. To see more stories like this one, change your Google preferences here.
Cool Picks
