AI is more likely than humans to form biases when hiring

A new study indicates that AI models, such as ChatGPT and Claude, may develop their own biases when evaluating job applicants. Researchers from Princeton University and the University of Chicago conducted a hiring simulation for 20 jobs, categorizing applicants into four fictional ethnic groups. Despite all candidates being equally likely to succeed, the models began segregating candidates based on early observations, leading to clear discrimination among groups.