A study by researchers at the Oxford Internet Institute has identified bias in large language models (LLMs) by asking ChatGPT ...