New research suggests that large language models, like ChatGPT, can develop their own biases from experience and stereotype job applicants more than humans do. In a simulated hiring game, AI models quickly started segregating candidates from different ethnic groups into different jobs based on early hiring outcomes.