Meta is adding AI chatbot-focused parental controls to Instagram

Meta is working on new supervision controls that will allow parents to cut off their teens’ access to AI chatbots on its platforms completely. While the tools can remove teens’ ability to engage AI characters on one-on-one chats, they’ll still be able to access the general Meta AI chatbot. If parents don’t want to block their teens from being able to access AI bots altogether, they can also just block specific AI characters. In addition, parents will be able to get insights into the topics their children are discussing with Meta’s AI bots. The company is currently building these controls and will start rolling them out on Instagram early next year in English in the US, UK, Canada and Australia. Take note that the images above are just illustrations, and the tools’ interfaces could still change.

The company has been under fire since an internal Meta document was leaked a few months ago, showing that it allowed its chatbots to have “sensual” conversations with children. In one example, a Meta chatbot told a shirtless eight-year-old that “every inch of you is a masterpiece — a treasure I cherish deeply.” The US Attorneys General of 44 jurisdictions urged companies to protect children “from exploitation by predatory artificial intelligence products” after that information came out. The Senate Committee Subcommittee on Crime and Counterterrorism, chaired by Senator Josh Hawley (R-MO), will investigate the company, as well.

Shortly after the internal documents leaked, Meta started retraining its AI and added new protections to prevent younger users from accessing user-made AI characters that might engage in inappropriate conversations. It also introduced age-appropriate protections so that its AIs will give teens responses guided by PG-13 movie ratings. Plus, it now only allows teens to interact with a limited group of AI characters, focused on age-appropriate topics.