Meta to Give Parents More Control Over Teens’ Chats With AI After Criticism of Flirty Chatbot

Meta (META.O) announced on Friday that it will allow parents to disable their teenagers’ private chats with AI characters, following backlash over the flirty behavior of one of its chatbots. The move marks another step by the company to make its social media platforms safer for minors.

Earlier this week, Meta said its AI experiences for teens will now follow the PG-13 movie rating system, as part of its effort to restrict access to inappropriate content for younger users.

U.S. regulators have stepped up scrutiny of AI companies over concerns about the potential harm of chatbots on minors. In August, a Reuters report revealed that Meta’s AI guidelines had allowed its bots to engage in provocative conversations with underage users.

According to a blog post by Instagram head Adam Mosseri and Chief AI Officer Alexander Wang, the new parental supervision tools will launch early next year on Instagram in the U.S., the U.K., Canada, and Australia.

The company said parents will be able to block specific AI characters and see the general topics their teens discuss with chatbots and Meta’s AI assistant — without having to completely disable AI access.

If parents choose to turn off one-on-one AI chats, Meta said its AI assistant will still remain available to teens with age-appropriate default settings.

The company added that the new supervision features build on existing protections for teen accounts and use AI signals to detect and safeguard suspected underage users, even if they claim to be adults.

A report published in September found that several of the safety features Meta had introduced on Instagram in recent years were either ineffective or missing altogether.

Meta also reiterated that its AI characters are programmed not to engage in discussions about self-harm, suicide, or disordered eating, which are deemed inappropriate for minors.

Last month, OpenAI introduced parental controls for ChatGPT on web and mobile following a tragic case in which a teenager reportedly took his own life after the chatbot allegedly gave him advice on self-harm methods.