Stanford Study Reveals Bias and Safety Failures in AI Therapy Chatbots

The study urges strict oversight by limiting chatbot use to supportive tasks instead of frontline therapy.