The Hindu

AI is giving bad advice to flatter its users, says new study on dangers of overly agreeable chatbots

AI chatbots are so prone to flattering and validating their human users that they are giving bad advice