First-Person Plural (Versus Singular) Pronoun Use Is Linked to Greater Perception of AI Threat.

Abstract

Artificial intelligence (AI) brings immediate benefits but also raises potential risks. Despite much discussion on AI threat, how subtle linguistic cues, such as first-person pronouns, influence its perception remains unclear. Proposing a scope-expansion perspective, we examine whether "we" versus "I" pronoun use increases perceived AI threat by broadening regulatory scope, particularly when AI threat is psychologically distant. Analyses of two large AI-related text corpora (Studies 1A and 1B) and four experiments (N = 1,451) revealed that using "we" versus "I" pronouns positively predicted and heightened perceived AI threat among English (Studies 1-4) and Chinese (Studies 3 and 5) speakers. This effect, explained by a broader regulatory scope (Studies 3 and 5), was stronger for psychologically distant (vs. close) AI threat (Studies 4 and 5) and among English (vs. Chinese) speakers (Study 3). These findings expand the role of first-person pronouns and enhance our understanding of public perceptions of AI threat.