Predicting decision-makers' coping strategies in human-AI conflict based on the link between personality traits and cognitive AI appraisals: A quantitative study of U.S. employees using AI at work

Abstract

As powerful AI aids become increasingly integrated into workplace decision-making, it becomes crucial to understand how employees respond to their recommendations, particularly when those recommendations conflict with their own judgments. We draw on cognitive appraisal theory and conflict management framework to examine how personality traits relate to different employee-AI conflict coping strategies via distinct appraisals of AI aids. In a three-wave study of 827 U.S. employees who regularly use AI aids at work, we measured the Big Five personality traits, employees' appraisals of these aids as both a threat and an opportunity to themselves, and four different types of conflict coping strategies when their judgment differs from that of the AI aid – avoiding, overruling, integrating and complying with the AI aid. Structural equation modeling with mediation analysis revealed nuanced indirect pathways. Beyond employees' interpersonal (human-human) conflict coping strategies and their use of AI aids (in terms of frequency and duration), we identified the following key findings: neuroticism and agreeableness showed exactly the opposite patterns – specifically, neuroticism (agreeableness) showed greater (lower) AI avoiding and overruling via increased (reduced) threat appraisals and less (more) integrating and complying via reduced (increased) opportunity appraisals; conscientiousness and openness to experience exhibited the same pattern – lower AI avoiding and overruling through reduced threat appraisals. This study provides new insights into employees' responses to conflicting AI recommendations, highlighting implications for implementation and augmentation strategies that consider individual trait differences. • How personality traits shape the way employees manage conflicts with AI? • The Big-five traits relate to different employee-AI conflict coping strategies. • These strategies are mediated by distinct (threat vs opportunity) AI appraisals. • Findings enable personality-based interventions for optimal human-AI collaboration.

Ähnliche Arbeiten