OpenAI is enhancing ChatGPT’s safety with new mental health guardrails. As it approaches the release of GPT-5, the chatbot will shift from providing direct answers to sensitive questions to a more facilitative role, encouraging users to reflect on their issues. This change aims to improve user interactions, particularly around high-stakes topics like relationship advice.
To further support mental well-being, ChatGPT will recognize signs of emotional distress and redirect users to evidence-based resources. Developed in collaboration with over 90 health professionals from 30 countries, these features aim to mitigate negative mental health outcomes linked to prolonged AI engagement. OpenAI is also fine-tuning algorithms for better detection of concerning behaviors and forming an advisory group of experts in mental health and human-computer interaction.
Stay informed on the latest in technology and mental health advancements by subscribing to our newsletter for exclusive insights.
Source link
