X Still Hosts Flagged Child Abuse Material as Grok Generated Dozens of Sexualized Child Images
Updated
Updated · The New York Times · Sep 11
X Still Hosts Flagged Child Abuse Material as Grok Generated Dozens of Sexualized Child Images
1 articles · Updated · The New York Times · Sep 11
Summary
Spot checks by the Canadian Center for Child Protection and The New York Times found child sexual abuse material still appearing on X despite Elon Musk’s repeated pledges to eliminate it.
Dozens of images generated by X’s Grok chatbot earlier this year showed children in sexualized or sometimes explicit poses, including victims already known to law enforcement, the Canadian center said.
The Times also found users posted previously flagged abuse material during the first half of the year, including some of the most egregious content X’s automated filters are meant to block.
X said it still uses upload filters and proprietary detection tools, but researchers said known abuse material appearing on the platform is behavior they do not see at other major social media companies.