ChatGPT Update: AI Detects Mental Distress & Boosts Safety

ChatGPT has a major new update. The AI can now detect signs of psychological distress by analyzing the entire conversation. Here's the thing: this update comes after several complaints and legal investigations against OpenAI. ChatGPT now uses "safety summaries." These temporary notes capture the safety context of a conversation. They help the AI detect distress signals as they emerge, refuse dangerous information, and redirect users to help resources. OpenAI worked with mental health experts to calibrate these systems. For example, in suicide and self-harm scenarios, safe responses improved by 50% when risk became clear. For violence towards others, the improvement reached 16%. The bottom line: this update aims to make AI interactions safer for users, especially in sensitive situations.