Paper Number

ECIS2026-2514

Paper Type

SP

Abstract

This study investigates how the disclosed identity of a legal decision-maker (human, AI, or human–AI) shapes perceptions of fairness. We conducted an online experiment (N = 264), where participants evaluated the fairness of a legal decision while receiving different information about its source. Decisions attributed to AI judges were rated significantly less fair on procedural and interactional dimensions, whereas hybrid human–AI decisions were viewed similarly to human decisions. These results suggest that full automation undermines perceived fairness, while human oversight remains critical. Our findings offer implications for designing and governing AI-assisted legal decision-making, indicating that hybrid models may preserve trust while enhancing efficiency.

Share

COinS
 
Jun 14th, 12:00 AM

Ai, Human, And Hybrid Judges: Effects On Fairness Perceptions In Legal Decisions

This study investigates how the disclosed identity of a legal decision-maker (human, AI, or human–AI) shapes perceptions of fairness. We conducted an online experiment (N = 264), where participants evaluated the fairness of a legal decision while receiving different information about its source. Decisions attributed to AI judges were rated significantly less fair on procedural and interactional dimensions, whereas hybrid human–AI decisions were viewed similarly to human decisions. These results suggest that full automation undermines perceived fairness, while human oversight remains critical. Our findings offer implications for designing and governing AI-assisted legal decision-making, indicating that hybrid models may preserve trust while enhancing efficiency.