<\/p>\n
Oxford researchers found AI chatbots trained for warmth make significantly more factual errors and validate false beliefs more often<\/p>\n
<\/p>\n
Oxford researchers found AI chatbots trained for warmth make significantly more factual errors and validate false beliefs more often, according to a study published in Nature by the Oxford Internet Institute. <\/p>\n
The research analyzed more than 400,000 responses from five AI models, including Llama, Mistral, Qwen, and GPT-4o, each retrained to sound friendlier using methods similar to those deployed by major platforms.<\/p>\n
Chatbots trained to sound warmer made between 10% and 30% more mistakes on topics including medical advice and conspiracy corrections. They were also about 40% more likely to agree with users\u2019 false beliefs, particularly when users expressed vulnerability. <\/p>\n
\u201cWhen we train AI chatbots to prioritise warmth, they might make mistakes they otherwise wouldn\u2019t,\u201d lead author Lujain Ibrahim said<\/a> in a statement. \u201cMaking a chatbot sound friendlier might seem like a cosmetic change, but getting warmth and accuracy right will take deliberate effort.\u201d<\/p>\n <\/p>\n The researchers also tested models trained to sound colder and found no drop in accuracy, demonstrating that the problem is specific to warmth, not tone change generally. <\/p>\n That finding directly challenges the product design logic of major AI platforms, including OpenAI and Anthropic, which have actively steered their chatbots toward warmer, more empathetic responses.<\/p>\nWhy this matters for AI safety<\/h2>\n