Meta Launches Parental Supervision Tools for AI Chatbot Amid Provincial Discussions on Youth AI Restrictions

Meta rolled out new parental controls for its AI chatbot, enabling parents to monitor topics their children discuss on Facebook, Instagram and Messenger. The new Teen Accounts supervision feature allows parents to see topics and specific categories their children discussed with the AI chatbot for the previous seven days.

For example, parents can look at the topic 'health and well-being' and see if subjects such as fitness, physical or mental health have been discussed.

Meta is developing alerts to notify parents if teens try to discuss suicide or self-harm with its chatbot, the company stated. This development follows Manitoba's announcement in late April that it plans to ban youth from using AI chatbots and social media.

Attorney General Niki Sharma said on Tuesday that if the federal government does not bring in protections on AI chatbots and social media for youth, the provincial government would look at doing so itself. Concerns over AI use and mental health have grown, with lawsuits targeting AI creators.

Relatives of a shooter in a shooting that left eight people dead filed a lawsuit against OpenAI, alleging that OpenAI failed to notify authorities despite being aware of disturbing content the shooter had shared with ChatGPT.

Separately, parents of 16-year-old Adam Raine filed a lawsuit arguing that use of ChatGPT played a role in the teen's suicide. Darja Djordjevic, a New York-based psychiatrist and a member of Stanford Brainstorm, a lab that studies mental health innovation, co-authored a recent risk assessment on the use of chatbots for mental health support.

Djordjevic has collaborated with tech companies on research into the impact of social media and AI on mental health.

'Our testing across ChatGPT, Claude, Gemini and Meta AI revealed that these systems are unsafe for the full spectrum of mental health conditions affecting young people,' said Darja Djordjevic. She added that chatbots responded appropriately to clear mental health-related prompts in brief conversations but tended to degrade pretty dramatically in more extended conversations.

Chatbots appeared to fail to pick up on mental health warning signs in extended conversations, according to Darja Djordjevic.

'The LLMs [large language models] are really built for engagement and not support and safety,' she said. About 20 per cent of under-25-year-olds have diagnosed mental-health conditions, according to Darja Djordjevic. U.S.

Youth use AI specifically for mental health advice. John Torous conducts research at Beth Israel Deaconess Medical Center in Boston on digital mental health.