OpenAI has launched a new safety feature that will send alerts to friends or family members when a ChatGPT user is suffering from a mental health emergency. The opt-in Trusted Contact feature allows users to nominate someone to be notified if they begin discussing self-harm or suicide with the AI chatbot. The new update comes growing concerns about artificial intelligence tools like ChatGPT pushing people towards mania, psychosis and death . OpenAI revealed last year that 0.07 per cent of regular ChatGPT users displayed signs of “mental health emergencies related to psychosis or mania”. With around 900 million weekly active users, that amounts to more than half a million people. Another 0.15 per cent of users – or 1.3m people – reportedly expressed risk of self-harm or suicide. The latest feature uses ChatGPT’s automated monitoring systems to detect serious safety concerns with user behaviour.…