[Goldman Sachs Analyst Reveals Murder Plan to ChatGPT, OpenAI Reports to Authorities] According to a report by Jinse Finance, a 25-year-old man in the United States was detected and reported to the Federal Bureau of Investigation (FBI) by OpenAI after he disclosed detailed plans to rape and murder his ex-girlfriend to the AI chatbot ChatGPT during work hours. AI companies often use multi-label harmful content classification models to accurately identify dangerous content in categories such as violence, hate, and sexual assault. They also assess risk levels based on emotional intensity and tone intent. According to OpenAI's company policy, when backend monitoring detects that a user may be planning to harm others, the content undergoes manual review. If deemed highly dangerous, the company may provide the conversation content to law enforcement as investigative leads. Court documents reveal that the man, named Darren Zhou, appeared to be of Asian descent. At the time of the incident, he was a Goldman Sachs analyst in West Palm Beach, Florida. FBI agents linked the AI conversation content to messages received by the woman, determining that Darren posed a sufficient threat to her personal safety. He was arrested in May. Darren was released on bail after posting a $100,000 bond. Goldman Sachs terminated his employment in June after learning of the charges. Darren was charged with offenses including making threats to kill or harm others and engaging in severe stalking with explicit threats to personal safety. On August 13, he pleaded guilty in court and was sentenced to five years of probation. The case has also sparked industry discussions about the boundaries of AI chat privacy: how platforms should define the scope of monitoring private user conversations and under what circumstances they have the right to submit user conversation content to law enforcement.