ABSTRACT:
Integrating artificial intelligence (AI) into organizations introduces challenges, particularly when AI-generated recommendations conflict with human judgment. Drawing on the theory of cooperation and competition and cognitive dissonance theory, we examine how human-AI conflict (low vs. high) and human-AI goal interdependence (cooperation vs. competition) shape cognitive dissonance and AI aversion across critical and non-critical decisions. Using an experiment with 432 employees, results show that higher human-AI conflict increases cognitive dissonance, which in turn reduces epistemic and social motivation and heightens AI aversion. The impact of human-AI conflict on cognitive dissonance intensifies when AI is perceived as a competitor and when decisions are critical. Moreover, in critical decisions, perceiving AI as a competitor amplifies the dissonance arising from human-AI conflict, whereas in non-critical decisions, cooperative and competitive perceptions of AI produce similar dissonance levels. These findings underscore the importance of aligning AI deployment with users’ relational perceptions and decision criticality.
Key words and phrases: Artificial intelligence, human-AI conflict, human-AI goal interdependence, cognitive dissonance, epistemic motivation, social motivation, AI aversion, decision criticality