Meta Platforms Users Report Being Wrongfully Locked Out After Purported AI Moderation Flags Accounts for Child Exploitation Content

Starting around July 2, 2023, multiple Meta platform users across the US received notifications that their Instagram, Facebook, and WhatsApp accounts had been disabled for allegedly violating community standards on child sexual exploitation, abuse and nudity. The affected users, including a pediatric ICU nurse from Azusa named Natalia and a Long Beach resident named Marty Macias, reported that the accusations were completely unfounded. When Natalia's account was disabled, her boyfriend Gavino Cervantes' professional Instagram account for his general contracting business was also disabled because their emails were connected. Cervantes reported losing approximately five potential customers who had been messaging him for quotes on the day his business account was taken down. The affected users submitted appeals but had not received responses from Meta. All Meta-owned platforms (Instagram, Facebook, WhatsApp) were disabled for these users, causing them to lose years of personal memories and business communications. Users suspected that an AI program had wrongfully flagged their accounts, and USC Communications Professor Karen North commented that this appeared to be an issue with over-reliance on early AI rollout. Meta had not responded to ABC7's request for comment, and users reported having no access to customer service support to resolve the issue.