Artificial Intelligence (AI) is increasingly automating expert judgments across diverse domains. However, the practical dynamics of adaptation among diverse stakeholders remain underexplored. We investigated the Korea Baseball Organization’s adoption of the Automated Ball-Strike System (ABS), the first league-wide deployment of an AI adjudicator. Interviews with 38 stakeholders—umpires, players, coaches, and fans—revealed that adoption was driven by demands for fairness and frustration with human limitations, and was viewed as an inevitable trajectory. Acceptance depended less on accuracy than on verifiable consistency, which reduced interpersonal conflict by shifting judgment to technology. However, adaptive burdens were redistributed: players faced pressure to recalibrate strategies for survival, while umpires grappled with diminished authority. Systemic legitimacy hinged on procedural transparency and visible feedback mechanisms. Based on these findings, we propose governance principles emphasizing transparency and adaptive role reconfiguration for sustainable human-AI coexistence.
ACM CHI Conference on Human Factors in Computing Systems