ChatGPT Gets Safety Overhaul

OpenAI has introduced new safety measures for ChatGPT, including a system that routes sensitive conversations to more advanced reasoning models like GPT-5. This routing system is designed to identify users who may be in distress and provide more careful and appropriate responses.

Parental controls are also being rolled out, allowing parents to link their accounts with their teenagers' accounts. These controls enable parents to manage how ChatGPT responds to their teen, set age-appropriate behaviour rules, disable memory and chat history, and receive alerts if the system detects signs of distress. Parents and teens must consent to linking accounts, and parents won't have access to their teen's chats unless a serious safety risk is detected.

These safety enhancements are part of a broader initiative to improve AI safety and address concerns about the potential for harm, especially to younger users. OpenAI is collaborating with mental health professionals and experts to guide the development and implementation of these features. Additional measures include an age-prediction system to identify underage users and direct them to a filtered version of ChatGPT with restricted content and functionality.