A scoping review maps five categories of mental health harms tied to LLM-based chatbots.
Two Harvard-affiliated physicians warn consumer chatbots are quietly building a parallel medical system.