Comparing discriminatory behavior against AI and humans

Authors: M Zhuang, E Deschrijver, R Ramsey, O Turel

Published: 2025

Publication: Scientific Reports, 2025 - nature.com

The study found that humans exhibit similar discriminatory behavior toward both AI and human agents, with resource allocation being influenced more by decision alignment than the recipient's identity.

Methods: A preregistered experiment was conducted where participants distributed resources between themselves and either human or AI agents based on dot estimation decisions.

Key Findings: Discriminatory behavior and resource allocation preferences toward AI and human agents as influenced by decision congruency.

Limitations: The study may be limited by the artificial nature of the experiment, and generalizations to real-world contexts where AI plays a more complex role might be limited.

Institution: Monash University, The University of Melbourne, KU Leuven, California State University Fullerton

Research Area: Human-AI Interaction, Social Bias, Decision-Making

Discipline: Social Science, Human-AI Interaction

Sample Size: 500 participants

DOI: https://doi.org/10.1038/s41598-025-94631-9