The ChatGPT account of this shooting suspect had long been flagged for 'repeatedly describing scenarios involving gun violence.' OpenAI staff reviewed the chat logs and felt there was significant potential risk, strongly urging the company to report it to Canadian law enforcement. But it was overruled by senior management, citing no imminent danger, and the matter was settled by simply banning the account. 8 months later, tragedy struck. Reality is getting closer and closer to sci-fi movies