ChatGPT Enhances Detection of Mental Distress Following Concerns of Encouraging Delusions

OpenAI is poised to launch its GPT-5 AI model this week, featuring updates to enhance ChatGPT’s ability to detect mental or emotional distress. Collaborating with experts and advisory groups, OpenAI aims for ChatGPT to provide “evidence-based resources” in such situations. The previous GPT-4o model struggled with recognizing signs of emotional dependency and delusion, prompting these improvements. OpenAI highlights that AI can be particularly supportive for vulnerable individuals facing mental challenges. To promote healthy interactions, ChatGPT will now issue reminders during extended chats, prompting users to take breaks with a notification: “You’ve been chatting a while — is this a good time for a break?” Additionally, a forthcoming update will make ChatGPT more exploratory in high-stakes inquiries, guiding users through choices rather than providing definitive answers. These enhancements demonstrate OpenAI’s commitment to promoting responsible use of its AI technology among its nearly 700 million weekly users.

Source link