Study: Chatbots are getting better at identifying suicide risk

https://www.axios.com/2026/08/31/chatbots-suicide-risks-identification

Independent coverage of Transluce's mental-health evaluation: "A new independent evaluation finds that leading AI models are far less likely than earlier versions to explicitly encourage suicide or reinforce delusions" and "the study also found most models too willing to assist with suicide-related creative writing"; names Transluce as a San Francisco nonprofit and quotes its chief scientist. Non-self confirmed source for transluce.02.