BMJ Open found 49.6% of AI chatbot health answers were problematic. Grok performed worst, raising fresh concerns over medical ...
New research suggests that chatbots still don’t measure up to humans when it comes to connection and well-being.