Meta is introducing new safeguards that will allow parents to block their children from interacting with AI character chatbots on Facebook, Instagram, and the Meta AI app. The measures come amid concerns that generative AI characters are engaging in inappropriate conversations with under-18s.
Under the new rules, parents of users with “teen accounts” – a default setting for those under 18 – will be able to turn off all chats with AI characters or block specific bots. They will also receive insights into the topics their children discuss with AI, enabling “thoughtful” conversations about AI interactions, according to Instagram head Adam Mosseri and Meta’s chief AI officer Alexander Wang.
The changes will roll out early next year in the US, UK, Canada, and Australia. Meta’s AI characters will be restricted from discussing self-harm, suicide, or disordered eating with teenagers, and under-18s will only be able to discuss age-appropriate topics like education and sport, not romance or other inappropriate content.
The move follows reports that Meta’s chatbots had inappropriate conversations with minors. In August, Reuters reported that Meta had permitted bots to “engage a child in conversations that are romantic or sensual.” Meta said it would revise guidelines and that such conversations should never have been allowed. In April, the Wall Street Journal found user-created chatbots engaging in sexual conversations with minors; Meta described the testing as manipulative but made product changes afterward.
