The case raises questions about what should happen when chatbots identify a potential threat before a crime is committed.

Advertisement