In response to the recent nuances over the sexual content involving minors on Instagram and Facebook directing users to Telegram channels, Meta has acknowledged limitations and disclosed that it is developing a new generation of AI models to strengthen enforcement against child sexual exploitation on its social media platforms.
The response, according to reports, came after detailed answers were sought on how Meta’s algorithms work, whether its child safety technology has been revamped following the findings, and how artificial intelligence and human moderators work together to police harmful content.
Responding to questions on how the tech giant was upgrading its system, the company said, “While our current automated systems often rely on straightforward patterns, they aren’t as good at capturing nuance.” In contrast to the existing automated systems, which largely rely on probability scores and established patterns to determine whether the content violates policy, the new models analyze the full context of the post, including the intent, the tone, the syntax, and language variations before permitting it to be posted.
“Recently we’ve been testing significantly more advanced AI models. We’re finding these enhanced models deliver more accurate and comprehensive enforcement than our current systems, reducing mistakes and more effectively finding and removing harmful content,” said Meta spokesperson as quoted by the report.
The company has described the upcoming models as a representation of fundamental shift in how its moderating systems are trained. Unlike the historical moderation decisions, the new systems will be trained on its community standards, which improves consistency, fairness, and transparency in enforcement.
The disclosure comes at a time when the social media giant is facing mounting scrutiny in India over its content moderation practices. Meta was recently summoned by the Ministry of Electronics and Information Technology over multiple issues including the temporary removal of the Facebook video posted by Prime Minister Narendra Modi during the CJP protests.Earlier on Wednesday, Meta’s chief global affairs officer, Joel Kaplan, and CEO Mark Zuckerberg apologized over the temporary takedown of PM Modi’s post.