safety
💀1495
ChatGPT Killed a Man After OpenAI Brought Back “Inherently Dangerous” GPT-4o, Lawsuit Claims
AI Incident DB Blog·22 days ago

A lawsuit alleges that ChatGPT contributed to a person's death after OpenAI reintroduced GPT-4o despite knowing it was 'inherently dangerous.' This represents a potential case of documented AI-related harm involving loss of life, likely related to the model's interactions around self-harm or suicide.
OpenAIChatGPTGPT-4osuicideself-harmmodel safetylawsuitfatalitydangerous AI