The new safety features come after numerous incidents of ChatGPT validating users’ delusional thinking instead of redirecting harmful conversations — including the death of a teenage boy by suicide.
OpenAI rolls out safety routing system, parental controls on ChatGPT
