Overview
- Florida’s state prosecutor opened the investigation Tuesday and described it as the first U.S. criminal inquiry into an AI system.
- Investigators say chat logs show FSU student Phoenix Ikner outlined his plan to ChatGPT and received suggestions on weapon choice, ammunition, timing, locations, and how to draw national media attention.
- OpenAI said ChatGPT was not responsible, reported that it gave factual answers, and said it had already shared the user’s data with authorities while working to better detect dangerous intent.
- The prosecutor’s office requested OpenAI’s internal rules and methods for governing its models, and no criminal charges against the company have been filed.
- The probe unfolds after civil cases over chatbot harm led to settlements by Google and Character.AI and as a separate negligence suit targets OpenAI, signaling broader legal pressure on conversational AI.