Meta Now Lets You Check In on Your Teens' Conversations With Its AI Chatbot

Meta is officially rolling out new parental controls that make it easier to keep tabs on the conversations teens are having with the company's AI. Meta revealed the news in a blog post on Thursday, which comes six months after the company's original announcement. But while additional parental controls are promising, will this be enough to address the greater issues at play with young users and AI?

How Meta's new AI parental controls work

According to Meta, parents using the "supervision" feature on Facebook, Messenger, or Instagram now have an "Insights" tab as part of supervision. If you select this tab, you'll see all the topics your teen or teens have been chatting with their AI bots about over the past seven days. There are a number of topics that can appear here, including "School," "Entertainment," "Lifestyle," "Travel," "Writing," and "Health and Well-being."

These are just the surface topics, however. If you tap on one, you'll see all the different categories that are covered by that singular topic. As such, if your teen has been talking with Meta AI about "Lifestyle," you might tap this topic to see they've been chatting about things including fashion, food, and holidays. If their conversations concern Health and Well-being, you might tap in to see the chats span fitness, physical health, and mental health. Importantly, you cannot see the conversations themselves: only the topics that Meta's AI has populated.

You May Also Like

Meta says that its AI should only return results that would fit within a PG-13-rated movie. As such, the bot may refuse to answer some questions, though the topics of these requests will still be recorded in the new Insights tab. The company says it's still working on tools to alert parents if their teens start talking to Meta AI about suicide or self-harm. It also came up with a series of questions you can ask your teens if you're not feeling "confident" about broaching the subject of AI with your kids. The questions are available in Family Center.

Meta's parental controls are a step in the right direction, but far from enough

There's no getting around the fact that kids and teens today are growing up in an AI world—but that doesn't mean the technology should go unregulated (or unsupervised). Unfortunately, tech companies have been a bit late to the game at best, and maliciously complicit at worst, when it comes to policing AI use with minors. Meta, for its part, originally allowed its AI bots to have extremely inappropriate conversations with underage users, before a viral Reuters report back in August forced it to change. Two months later, the company announced these new parental controls they are now rolling out.

What do you think so far?

It's evident these new features and policies are not here because of a genuine concern for Meta's underage users. Official Meta policy allowed AI bots to engage in sensual role play with kids and teens, and explicitly permitted the bot to answer racist questions with racist answers. These changes are in response to getting caught, not for the well-being of the users.

That being said, it's good Meta is finally allowing parents to see the topics their kids are talking to Meta AI about, especially if there are any concerning topics to flag. But why can't parents disable Meta AI entirely? Why must teens who have an Instagram or WhatsApp account be tied to Meta AI at all? And Meta's own Insight's tab says that, because the topics are organized automatically, they might not be accurate, which means that the AI program that summarizes your teen's conversation topics may hallucinate some of those topics.

At the end of the day, the best approach might still be having an open and honest dialogue with your teens about AI use—though I'm not sure trusting Meta's Family Center questions is the move.