Stanford University researchers have found that AI chatbots like ChatGPT can provide "dangerous or inappropriate" responses to users experiencing mental health crises, potentially escalating episodes of suicidal ideation, mania and psychosis.
Artificial Intelligence