How human-AI feedback loops alter human perceptual, emotional and social judgements
Moshe Glickman, Tali Sharot
Abstract
Artificial intelligence (AI) technologies are rapidly advancing, enhancing human capabilities across various domains spanning from finance to medicine. Despite their numerous advantages, AI systems can exhibit biases in judgments ranging from perception to emotion. Here, in a series of experiments (N=1,201), we reveal a feedback loop where human-AI interactions alter processes underlying human perceptual, emotional and social judgements, subsequently amplifying biases in humans. This amplification is significantly greater than observed in interactions between humans, due both to the tendency of AI systems to amplify biases and to how humans perceive AI systems. Participants are often unaware of the extent of the AI’s influence, rendering them more susceptible to it. These findings reveal a mechanism wherein AI systems amplify human biases, which are further internalized by humans during human-AI interactions, triggering a snowball effect where small errors in judgment escalate into much larger ones.