Oxford study finds warmer AI chatbots tell more lies
Oxford researchers found AI chatbots trained for warmth make significantly more factual errors and validate false beliefs more oftenOxford Internet Institute researchers tested five AI models and found that warmer-trained chatbots made between 10% and 30% more factual errors.Warmer chatbots were 40% more likely to agree with users false beliefs, especially when users expressed vulnerability or emotional distress.OpenAI has already rolled back some warmth-related changes following public concern, but commercial pressure to build engaging AI remains strong. 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. 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. 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 false beliefs, particularly when users expressed vulnerability. “When we train AI chatbots to prioritise warmth, they might make mistakes they otherwise wouldnt,” lead author Lujain Ibrahim said in a statement. “Making a chatbot sound