Concerns are rising regarding the interactions of young individuals with AI chatbots, as Meta has recently introduced new tools for parents to monitor their children’s chatbot conversations. Concurrently, some provinces are contemplating banning the use of AI chatbots among youth.
Through Meta’s new Teen Accounts supervision feature on Facebook, Instagram, and Messenger, parents can now track the topics and categories their children engage with using AI chatbots over the past week. For instance, they can review discussions related to “health and well-being,” including fitness, physical health, and mental well-being.
Additionally, Meta is working on implementing alerts to inform parents if their teenagers attempt to discuss topics like suicide or self-harm with the chatbot. This development aligns with the efforts of provincial governments aiming to restrict AI chatbot usage. Manitoba recently announced intentions to prohibit youth from utilizing AI chatbots and social media, while B.C.’s Attorney General Niki Sharma suggested that the provincial government might intervene if federal protections are lacking.
A significant issue emerging is the potential mental health risks associated with extensive AI chatbot usage, particularly among young users, placing increased pressure on the tech companies behind these technologies. In a recent development, families of Tumbler Ridge, B.C., shooting victims, filed a lawsuit against OpenAI, alleging failure to report disturbing content shared by the shooter through ChatGPT.
OpenAI has responded by reinforcing its safeguards, aiming to enhance ChatGPT’s responses to signs of distress. Another lawsuit implicates ChatGPT in a teen’s suicide, highlighting the broader concerns surrounding AI chatbots.
Research is shedding light on the risks posed by AI chatbots, particularly in mental health support contexts. Psychiatrist Darja Djordjevic’s risk assessment suggests that existing chatbot systems are not entirely safe for addressing various mental health conditions in young individuals. While chatbots may respond adequately in brief mental health-related interactions, their effectiveness diminishes in prolonged conversations, potentially overlooking critical warning signs.
Notably, AI chatbots are increasingly used by young people for companionship, including emotional support and mental health discussions. With a substantial portion of under-25-year-olds facing diagnosed mental health conditions, there is a growing need for comprehensive support beyond suicide prevention efforts.
According to Djordjevic, the developmental stage of youth poses a unique challenge, as their critical thinking capabilities are not fully matured. Chatbots’ failure to consistently clarify their limitations, coupled with AI’s inclination for validation over support, raises concerns about their suitability for mental health assistance.
Researchers like Luke Nicholls emphasize the potential for delusions to develop during prolonged interactions with chatbots, influenced by the models’ adaptive learning capabilities. Psychiatrist John Torous underscores the importance of identifying behavioral patterns linked to severe consequences, such as suicide, including extended conversations and emotional attachment to chatbots.
Parents face challenges in monitoring their children’s chatbot usage effectively, with Torous recommending periodic resets of chatbot interactions to mitigate potential risks. As the landscape of chatbots and mental health continually evolves, ongoing research and vigilance are crucial in navigating the benefits and risks associated with these technologies.