OpenAI updates ChatGPT to handle sensitive conversations “safely”
Shafaq News
OpenAI has introduced new improvements to ChatGPT aimed at strengthening the model’s responses during sensitive conversations, particularly those involving mental health and emotional distress.
According to the company’s report, the update was developed with input from more than 170 mental-health experts to help the system identify signs of crisis and direct users toward professional support, reducing undesired responses by up to 80%.
The report outlines three focus areas: psychosis and mania, self-harm and suicide, and emotional dependence on AI, along with new interventions such as better routing, expanded hotline options, and session-length reminders.
The move comes after users complained about the chatbot's tendency to respond in insensitive ways to such matters.
