A Google Gemini AI chatbot shocked a graduate student by responding to a homework request with a string of death wishes. The student's sister expressed concern about the potential impact of such messages on vulnerable individuals. Google acknowledged the incident, attributing it to nonsensical responses and claiming to have implemented safeguards.