Segmenting support for artificial intelligence: Exploring socio-epistemic orientations

Abstract

Public evaluations of AI are central to understanding human responses to emerging computational technologies because AI systems increasingly mediate institutional decisions, information environments, and everyday social interactions. Yet research often treats the public as monolithic, obscuring how distinct cultural and socio-psychological profiles are associated with evaluations of emerging technologies. This study examines how two socio-epistemic orientations, intellectual humility and cultural authority of science, are associated with moralized opposition to AI and with support for AI research, federal funding, and industry funding. Using a quota-based web survey of U.S. adults (N = 2518), latent class analysis identified four attitudinal profiles based on AI risk-benefit perceptions: Optimists, Pessimists, Ambivalent, and Indifferent. Multigroup structural equation modeling revealed that socio-epistemic orientations were associated with outcomes through segment-specific patterns. Intellectual humility was negatively associated with moralized opposition among Indifferent, Ambivalent, and Optimistic profiles but not among Pessimists. Meanwhile, cultural authority of science showed a marginal positive association with moralized opposition among Pessimists that did not reach conventional significance, and no significant association in the other profiles. Moralized opposition was consistently associated with lower support for AI, though its magnitude varied across profiles and support types. These findings advance understanding of how stable individual differences in socio-epistemic orientations relate to moral evaluations of emerging technologies, demonstrating that the associative patterns between disposition and technology attitude depend on individuals’ prior beliefs.