OpenAI Adds Safeguards, Open-Sources AI, Faces Lawsuit

OpenAI has implemented new mental health safeguards in ChatGPT after reports of the chatbot occasionally reinforcing harmful or delusional behavior, including rare contributions to mental health crises. Updates include break reminders and a change in how ChatGPT handles high-stakes personal questions, focusing on reflective guidance rather than direct advice. OpenAI acknowledges previous shortcomings in detecting emotional distress and is now working with both mental health and human-computer interaction experts to improve safety protocols. ChatGPT's user base has surged to nearly 700 million weekly users, amplifying its societal impact and the importance of responsible AI use. OpenAI has also released open-weight AI models for commercial use, marking a shift toward open sourcing. Meanwhile, the company faces ongoing privacy concerns as it negotiates with The New York Times over access to user chat logs in an active lawsuit.