Abstract
<title>Abstract</title> <p> <bold>Background:</bold> Artificial intelligence (AI)-based agents increasingly participate in decisions that affect human experiences, raising a social-cognitive question about how people explain unfavorable outcomes after they occur. Less is known about how recipients explain and accept unfavorable outcomes after they have already been allocated. We examined whether agent type shapes outcome acceptance through upward counterfactual thinking and intentionality attribution, two forms of post-outcome causal interpretation, and whether these pathways depend on personal stake, conceptualized as the self-relevance and personal consequence of the outcome. <bold>Methods:</bold> Two scenario experiments were conducted. Study 1 used a food-delivery task allocation scenario to compare outcomes produced by artificial and human agents and test the parallel mediating roles of upward counterfactual thinking and intentionality attribution ( <italic>N</italic> = 472). Study 2 used an internal training quota allocation scenario and a 2 × 2 between-subjects design manipulating agent type and personal stake ( <italic>N</italic> = 360). It tested whether personal stake moderated the indirect pathways from agent type to outcome acceptance. <bold>Results:</bold> Across both studies, outcomes produced by artificial agents were accepted more readily than equivalent outcomes produced by human agents. Human agents elicited stronger upward counterfactual thinking and stronger intentionality attribution, and both processes were associated with lower outcome acceptance. Study 2 showed that differences between artificial and human agents were weak or nonsignificant under low personal stake but stronger under high personal stake, when the unfavorable outcome affected career-development resources more directly. <bold>Conclusions:</bold> Agent type influenced outcome acceptance partly through differences in upward counterfactual thinking and intentionality attribution. Personal stake functioned as a boundary condition by strengthening these explanatory responses when unfavorable outcomes carried direct personal consequences. The results show that how people represent the responsible agent shapes post-outcome cognition after an unfavorable outcome. </p>