Anthropomorphism and emotional reactions to human–AI collaboration.

As artificial intelligence (AI) increasingly enters human professional domains, understanding emotional reactions to its involvement becomes critical.We introduce a novel experimental paradigm that distinguishes between AI involvement in "Thinker" and "Doer" roles to examine how the role AI occupies shapes affective and evaluative responses to human-AI collaboration.Across four countries (N = 1,345), including the United States, the United Kingdom, India, and Colombia, participants evaluated scenarios in which professional jobs were completed under four conditions: human only, full AI automation, human as Thinker with AI as Doer, or AI as Thinker with human as Doer.Our results showed that AI involvement decreased positive affect and appreciation while increasing negative affect.We also found particularly negative emotional reactions to AI executing emotionally charged professions (e.g., artist).Individual differences in AI anthropomorphism consistently moderated these effects.These findings demonstrate that emotional reactions to AI depend on the role it plays and the domain in which it operates.This research highlights the importance of task characteristics and individual differences in understanding human-AI collaboration in professional settings.