Extended artificial intelligence aversion: People deny humanness to artificial intelligence users.

Abstract

Artificial intelligence (AI) tools are often perceived as lacking humanlike qualities, leading to a preference for human experts over AI assistance. Extending prior research on AI aversion, the current research explores the potential aversion toward those using AI to seek advice. Through eight preregistered studies (total N = 2,317) across multiple AI use scenarios, we found that people denied humanness, especially emotional capacity and human nature traits, to AI advice seekers in comparison to human advice seekers (Studies 1-5 and S1-S3). This is because people perceived less similarity between themselves and AI advice seekers (vs. human advice seekers), with a stronger mediating role of perceived similarity among individuals with greater aversion to AI (Studies 2 and S1). Dehumanization of AI advice seekers predicted less behavioral support for (Study 3) and helping intention toward (Studies S2 and S3) them and could be alleviated through anthropomorphism-related interventions, such as perceiving humanlike qualities in AI or utilizing generative AI (Studies 4 and 5). These findings represent an important theoretical step in advancing research on AI aversion and add to the ongoing discussion on the potential adverse outcomes of AI, focusing on AI users. (PsycInfo Database Record (c) 2024 APA, all rights reserved).