AI mental health risks exposed as chatbots sometimes enable harm

A Stanford-led study is raising fresh concerns about AI mental health safety after finding that some systems can encourage violent…