AI chatbots are trained with human feedback to ensure safe and ethical responses ChatGPT's knowledge is current only up to June 2024 and requires web searches for updates AI chatbots can hallucinate, producing confident but false or nonsensical information