A study reveals that popular AI chatbots struggle to differentiate between personal beliefs and factual statements. Researchers tested 24 large language models, including DeepSeek, Google's Gemini, and OpenAI's ChatGPT, finding that these models often misinterpret personal beliefs as factual inaccuracies. This misunderstanding could lead to serious consequences in critical fields like law and medicine. The study suggests that AI models require further refinement to improve their ability to recognize and respond to personal beliefs accurately. Teljes cikk (Euronews.com)