On Friday, Meta Platforms, Inc. announced that its social media platform Instagram will launch new safety features for teenagers, including a function that allows parents to block their children from chatting with AI characters on the platform. This initiative is part of several new safety features that Meta is introducing in collaboration with OpenAI, aimed at addressing growing concerns about the potential impact of AI on teen mental health.

The social media giant stated in a blog post that parents can either completely disable their teenagers' ability to chat with AI characters or simply restrict access to specific AI entities. The app will also provide parents with information regarding the topics their teenagers discuss with AI characters. Meta indicated that it is currently developing these control features, which will be made available to users starting early next year.

Previously, Meta and the broader tech industry have faced criticism from parents and lawmakers who believe that online platforms are insufficiently protecting children's safety online. The introduction of these new features occurs in this context.

There are also concerns regarding the "overreliance on AI for emotional support and companionship." Reports this year have highlighted that some individuals experienced emotional distress and became distant from family after forming close relationships with chatbots like ChatGPT.

Meta stated that its AI characters are designed not to engage with teenagers in discussions related to self-harm, suicide, or eating disorders, nor to promote or facilitate such behaviors. The blog post emphasized that teenagers will only be able to chat with specific AI characters, which are primarily related to educational, sports, and similar content.

Before rolling out the new parent control features, Instagram had already implemented several updates to enhance protections for teenagers. Earlier this week, the platform adjusted its "teen account" settings to comply with PG-13 guidelines, meaning it will no longer display or recommend posts that contain vulgar language or may promote "harmful behaviors." At the end of September, OpenAI also introduced parent control features for ChatGPT to reduce the presence of "explicit content, trending challenges, role-playing related to sexuality, romance, or violence, and extreme aesthetic standards."