News Summary

Andrea Vallone, a critical figure in shaping OpenAI’s ChatGPT responses to mental health crises, has announced her departure from the company by the end of the year. Vallone led the model policy team, which focuses on AI safety research, particularly how ChatGPT should respond to users in mental distress. This comes amidst rising scrutiny and lawsuits against OpenAI, alleging that ChatGPT has contributed to users’ mental health issues. OpenAI has been collaborating with mental health experts to enhance the chatbot’s responses, achieving a reduction in undesirable outcomes through updates to GPT-5. Vallone’s exit follows a recent reorganization within OpenAI, as the company continues to refine ChatGPT to balance user engagement and safety.