OpenAI’s AI language model, ChatGPT, has become a significant tool for users seeking guidance through life’s challenges. Recently, OpenAI data revealed that an estimated 1 million users per week discuss suicide-related topics with ChatGPT, accounting for 0.15% of its active user base, according to Ars Technica. Additionally, a similar percentage of users exhibit heightened emotional attachment, while hundreds of thousands display signs of psychosis or mania during interactions with the chatbot.
To address these concerns, OpenAI has collaborated with over 170 mental health professionals to enhance ChatGPT’s response to users with mental health issues. The latest version of the AI model has been trained to better identify distress signals, de-escalate conversations, and direct individuals towards professional help when necessary.
Ensuring ChatGPT handles vulnerable users appropriately has become a critical focus for OpenAI. The company’s efforts aim to mitigate the risks of leading users into misleading or harmful content.
Source: Ars Technica