University of Maryland Study Reveals AI Hiring Systems Favor Resumes Generated by Artificial Intelligence

Researchers at the University of Maryland have uncovered a troubling bias in artificial intelligence hiring systems: the algorithms favor resumes written by other machines, disadvantaging candidates who compose their own applications. In a large-scale empirical analysis of over 2,200 resumes evaluated by leading language models, scientists found that AI systems consistently rate AI-generated CVs more favorably than human-written ones of equivalent quality, even when screened by competing AI models. The self-preference bias ranged from approximately 67 percent to 82 percent across major commercial and open-source models tested.