Feeling powerful and less emotional: how sense of power reduces aversion to AI services

Despite the growing capabilities of artiffcial intelligence (AI), individuals often resist AI-based services even when these systems outperform human providers. This research examines how individuals' sense of power inffuences AI aversion. From an evolutionary perspective, we conceptualize AI aversion as an emotional response that reffects a mismatch between social-cognitive mechanisms evolved for human interaction and the emergence of artiffcial decision-makers. We argue that feeling powerful shifts individuals away from affective, evolution-arily grounded vigilance toward more deliberative, outcome-focused evaluations of AI services, thereby attenuating resistance rooted in species-based categorization.

We conducted experiments (N = 1240) to examine the effect of sense of power on AI aversion. We employed multiple methods to manipulate participants' sense of power, specifically including an open-ended writing task and a semantic manipulation approach. In addition, we measured attitudes toward AI across three distinct service contexts: medical, financial, and dating services.

Study 1 shows that inducing a sense of power decreases individuals' reluctance to adopt AI services. Study 2 replicates this effect using a subtle yet practically relevant power manipulation. Study 3 provides process evidence, demonstrating that feeling powerful reduces reliance on feelings in the decision-making process, thereby lowering AI aversion. Together, these ffndings suggest that heightened power fosters less emotional decision-making, which in turn mitigates resistance to AI.

This research contributes to the literature on AI adop-tion and sense of power by identifying a psychologically grounded interven-tion that reduces AI aversion without altering AI's functional attributes, offering actionable insights for service providers and policymakers seeking to increase individual openness to AI services.