Artificial intelligence tools like ChatGPT struggle to distinguish belief from fact, a new study has revealed.
A team from Stanford University in the US found that all major AI chatbots failed to consistently identify when a belief is false, making them more likely to hallucinate or spread misinformation.
