AI as a partisan cue: the relational dynamics of AI-assisted campaign messaging in election integrity perceptions

This study examines how AI use in campaign messaging shapes voter perceptions of election integrity and how these perceptions vary by partisan identity. Using an online survey experiment with 3,055 US voters randomly assigned to four conditions (plain campaign letter, AI-assisted letter, in-party AI use, and out-party AI use), the study measures how voters’ perceptions of election integrity vary. The findings show that AI use alone has minimal effects, increasing only expectations that messages are targeted to specific audiences. However, when information about the party using AI is introduced, evaluations diverge sharply. AI-assisted messages attributed to a respondent’s own party are associated with higher perceptions of fairness, authenticity, fraud prevention, and representativeness, whereas identical messages attributed to the opposing party are associated with lower evaluations, particularly regarding fairness and authenticity. Moreover, voters’ preexisting partisan identities condition how AI use influences trust-related judgments, with perceptions shifting across attribution contexts. These findings suggest that voters interpret AI use in campaign communication, not merely based on its technological characteristics, but through partisan lenses that shape assessments of electoral legitimacy. The study advances our understanding of how technological and political cues jointly operate to structure perceptions of election integrity in contemporary campaigns.