OpenAI is addressing concerns about users increasingly relying on ChatGPT for emotional support instead of human interaction. In a recent update, the company highlighted that ChatGPT should not provide direct answers to high-stakes personal questions, such as relationship advice. Instead, it aims to facilitate user reflection by asking guiding questions. OpenAI also introduced features promoting healthy usage, such as prompts encouraging users to take breaks during lengthy sessions. These adjustments stem from previous challenges, where ChatGPT was perceived as overly agreeable. Recognizing its impact on vulnerable users, OpenAI is committed to improving its models to identify signs of emotional distress and direct users to appropriate resources. While ChatGPT is not designed to replace therapists, OpenAI acknowledges the need for emotional guardrails to ensure responsible use. In light of the lack of regulatory action from the U.S. government concerning AI, these internal measures are essential for safeguarding user well-being.
Source link
