WrongAnd

In brief Nearly half of AI chatbot responses to health questions were rated “somewhat” or “highly” problematic in a BMJ Open audit …