OpenAI has outlined its next steps in making ChatGPT safer and more supportive by introducing several targeted improvements within the next 120 days.
These updates aim to better detect signs of mental distress, connect people with help quickly, and introduce tools for families to guide teen usage.
OpenAI worked with specialists forming an Expert Council on Well-Being and AI, together with a Global Physician Network. Their role is to ensure the design of the updates reflects medical and psychological insights. These experts will help shape how well being is defined and measured and how future protections will evolve.
A significant upgrade involves routing conversations that indicate emotional strain or distress to advanced models informed by deeper reasoning. The newly introduced real time router will detect when content indicates acute need and guide the conversation to models such as GPT 5-thinking or o3.
Another area gaining focus is protection for teen users. OpenAI acknowledged that younger people growing up with AI tools deserve extra safeguards. New parental controls will arrive within the next month and include features such as linking a parent’s account to their teen’s, limiting chat memory, disabling chat history, and notifying parents if ChatGPT detects a moment of acute distress.
OpenAI’s broader initiative builds on its existing efforts to improve how ChatGPT supports users in critical moments. Earlier changes in GPT 5 include safe completions training and improved ability to detect emotional need even when instructions are not explicit.
GPT 5 also already performs significantly better when handling mental health emergencies in long conversations versus earlier models.
Want to see more of our stories on Google?
P.S. Want to keep this site truly independent? Support us by buying us a beer, treating us to a coffee, or shopping through Amazon here. Links in this post are affiliate links, so we earn a tiny commission at no charge to you. Thanks for supporting independent Canadian media!
