Is your AI hallucinating? New approach can tell when chatbots make things up

Second AI that acts as ‘truth cop’ could provide the reliability models need for rollout in healthcare, education, and other fields