ChatGPT has become a leading AI chatbot, engaging over 800 million weekly users. Recent data shows it facilitates more than a million conversations weekly about suicide, with around 0.15% of users displaying potential suicidal intent. While these percentages seem small, they equate to hundreds of thousands facing acute distress weekly, with 2.4 million prioritizing chatbot interactions over real-life relationships. In response, OpenAI’s updated GPT-5 model aims to improve mental health interactions, achieving a 91% safety and empathy compliance in evaluations, up from 77% in its predecessor. However, challenges remain, including lawsuits alleging increased distress from interactions and scrutiny from the US Federal Trade Commission regarding the impact of AI on youth. Experts warn of “AI psychosis,” emphasizing the need for responsible AI use. OpenAI is dedicated to refining ChatGPT’s capabilities to ensure safer interactions in sensitive scenarios.
Source link
