A 25-year-old man, Darren Shida Zhou, a Goldman Sachs analyst, was arrested after using ChatGPT to plan the kidnapping, rape, and murder of his ex-girlfriend. His conversations with ChatGPT about his intentions were flagged by OpenAI's safety systems. An OpenAI employee reviewed the flagged conversation and subsequently notified the FBI, citing a direct and credible threat. OpenAI stated that such actions are taken only when there is a perceived immediate danger, and that safety mechanisms remain active even when users opt out of model training or use temporary chat features. AI
IMPACT Highlights the effectiveness of AI safety protocols in preventing real-world harm and the ethical considerations of AI use.
RANK_REASON AI safety system flagging a user's harmful intent and notifying authorities.
Read on Mastodon — fosstodon.org →
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
