Recent reports reveal AI chatbots are being tested on how they handle queries about violence and murder. Developers face a growing dilemma between protecting user privacy and preventing real-world harm. This tension is now making global headlines as governments and tech companies debate where the line should be drawn.